diff --git a/src/packagedcode/about.py b/src/packagedcode/about.py
index 3f33c6838db..9f9d4238861 100644
--- a/src/packagedcode/about.py
+++ b/src/packagedcode/about.py
@@ -47,7 +47,7 @@ class AboutFileHandler(models.DatafileHandler):
     documentation_url = 'https://aboutcode-toolkit.readthedocs.io/en/latest/specification.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         """
         Yield one or more Package manifest objects given a file ``location`` pointing to a
         package archive, manifest or similar.
@@ -71,6 +71,15 @@ def parse(cls, location):
 
         name = package_data.get('name')
         version = package_data.get('version')
+        if purl_only:
+            yield models.PackageData(
+                datasource_id=cls.datasource_id,
+                type=package_type,
+                namespace=package_ns,
+                name=name,
+                version=version,
+            )
+            return
 
         homepage_url = package_data.get('home_url') or package_data.get('homepage_url')
         download_url = package_data.get('download_url')
diff --git a/src/packagedcode/alpine.py b/src/packagedcode/alpine.py
index 59223cf72da..58ebdfcc0e3 100644
--- a/src/packagedcode/alpine.py
+++ b/src/packagedcode/alpine.py
@@ -63,11 +63,12 @@ class AlpineInstalledDatabaseHandler(models.DatafileHandler):
     description = 'Alpine Linux installed package database'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         yield from parse_alpine_installed_db(
             location=location,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -134,9 +135,10 @@ class AlpineApkbuildHandler(models.DatafileHandler):
     documentation_url = 'https://wiki.alpinelinux.org/wiki/APKBUILD_Reference'
 
     @classmethod
-    def parse(cls, location):
-        package_data = parse_apkbuild(location, strict=True)
-        cls.populate_license_fields(package_data)
+    def parse(cls, location, purl_only=False):
+        package_data = parse_apkbuild(location, strict=True, purl_only=purl_only)
+        if not purl_only:
+            cls.populate_license_fields(package_data)
         if package_data:
             yield package_data
 
@@ -165,7 +167,12 @@ def assign_package_to_resources(cls, package, resource, codebase, package_adder)
         )
 
 
-def parse_alpine_installed_db(location, datasource_id, package_type):
+def parse_alpine_installed_db(
+        location,
+        datasource_id,
+        package_type,
+        purl_only=False,
+    ):
     """
     Yield PackageData objects from an installed database file at `location`
     or None. Typically found at '/lib/apk/db/installed' in an Alpine
@@ -179,6 +186,7 @@ def parse_alpine_installed_db(location, datasource_id, package_type):
             package_fields=package_fields,
             datasource_id=datasource_id,
             package_type=package_type,
+            purl_only=purl_only,
         )
 
 
@@ -241,7 +249,7 @@ def get_alpine_installed_db_fields(location):
 ])
 
 
-def parse_apkbuild(location, strict=False):
+def parse_apkbuild(location, strict=False, purl_only=False):
     """
     Return a PackageData object from an APKBUILD file at ``location`` or None.
 
@@ -256,6 +264,7 @@ def parse_apkbuild(location, strict=False):
         datasource_id=AlpineApkbuildHandler.datasource_id,
         package_type=AlpineApkbuildHandler.default_package_type,
         strict=strict,
+        purl_only=purl_only,
     )
 
 
@@ -732,7 +741,13 @@ def fix_apkbuild(text):
     return text
 
 
-def parse_apkbuild_text(text, datasource_id, package_type, strict=False):
+def parse_apkbuild_text(
+        text,
+        datasource_id,
+        package_type,
+        strict=False,
+        purl_only=False
+    ):
     """
     Return a PackageData object from an APKBUILD text context or None. Only
     consider variables with a name listed in the ``names`` set.
@@ -761,7 +776,8 @@ def parse_apkbuild_text(text, datasource_id, package_type, strict=False):
     package = build_package_data(
         variables,
         datasource_id=datasource_id,
-        package_type=package_type
+        package_type=package_type,
+        purl_only=purl_only,
     )
 
     if package and unresolved:
@@ -800,7 +816,12 @@ def parse_pkginfo(location):
     raise NotImplementedError
 
 
-def build_package_data(package_fields, datasource_id, package_type):
+def build_package_data(
+        package_fields,
+        datasource_id,
+        package_type,
+        purl_only=False
+    ):
     """
     Return a PackageData object from a ``package_fields`` iterable of (name,
     value) tuples.
@@ -832,10 +853,17 @@ def build_package_data(package_fields, datasource_id, package_type):
         'type': package_type,
     }
     for name, value in package_fields:
-        handler = package_handlers_by_field_name.get(name)
+        handler = package_handlers_by_field_name_purl_only.get(name)
+        if not purl_only and not handler:
+            handler = package_handlers_by_field_name_others.get(name)
+
         if handler:
             try:
-                converted = handler(value, all_fields=all_fields, **converted_fields)
+                converted = handler(
+                    value,
+                    all_fields=all_fields,
+                    **converted_fields
+                )
             except:
                 raise Exception(*list(package_fields))
 
@@ -1199,11 +1227,11 @@ def source_handler(value, **kwargs):
 # mapping of:
 # - the package field one letter name in the installed db,
 # - an handler for that field
-package_handlers_by_field_name = {
+package_handlers_by_field_name_purl_only = {
 
-    ############################################################################
-    # per-package fields
-    ############################################################################
+    ###########################################################################
+    # per-package fields (only purl fields)
+    ###########################################################################
 
     # name of the package
     # For example: P:busybox
@@ -1218,6 +1246,22 @@ def source_handler(value, **kwargs):
     'V': build_name_value_str_handler('version'),
     'pkgver': apkbuild_version_handler,
 
+    # For example: D:scanelf so:libc.musl-x86_64.so.1
+    # For example: D:so:libc.musl-x86_64.so.1 so:libcrypto.so.1.1 so:libssl.so.1.1 so:libz.so.1
+    # Can occur more than once
+    # 'depend' in .PKGINFO and APKBUILD
+    # TODO: add other dependencies (e.g. makedepends)
+    'D': D_dependencies_handler,
+    'depend': D_dependencies_handler,
+}
+
+
+package_handlers_by_field_name_others = {
+
+    ###########################################################################
+    # per-package fields (other than purls)
+    ###########################################################################
+
     # For example: T:Size optimized toolbox of many common UNIX utilities
     # 'pkgdesc' in .PKGINFO and APKBUILD
     'T': build_name_value_str_handler('description'),
@@ -1272,14 +1316,6 @@ def source_handler(value, **kwargs):
     'c': c_git_commit_handler,
     'commit': c_git_commit_handler,
 
-    # For example: D:scanelf so:libc.musl-x86_64.so.1
-    # For example: D:so:libc.musl-x86_64.so.1 so:libcrypto.so.1.1 so:libssl.so.1.1 so:libz.so.1
-    # Can occur more than once
-    # 'depend' in .PKGINFO and APKBUILD
-    # TODO: add other dependencies (e.g. makedepends)
-    'D': D_dependencies_handler,
-    'depend': D_dependencies_handler,
-
     # For example: source="http://liba52.sourceforge.net/files/$pkgname-$pkgver.tar.gz
     #         automake.patch
     #         fix-globals-test-x86-pie.patch"
diff --git a/src/packagedcode/bower.py b/src/packagedcode/bower.py
index fe8c197cc87..d96f540a9e2 100644
--- a/src/packagedcode/bower.py
+++ b/src/packagedcode/bower.py
@@ -25,42 +25,13 @@ class BowerJsonHandler(models.DatafileHandler):
     documentation_url = 'https://bower.io'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with io.open(location, encoding='utf-8') as loc:
             package_data = json.load(loc)
 
         # note: having no name is not a problem for private packages. See #1514
         name = package_data.get('name')
-
-        description = package_data.get('description')
         version = package_data.get('version')
-        extracted_license_statement = package_data.get('license')
-        keywords = package_data.get('keywords') or []
-
-        parties = []
-
-        authors = package_data.get('authors') or []
-        for author in authors:
-            if isinstance(author, dict):
-                name = author.get('name')
-                email = author.get('email')
-                url = author.get('homepage')
-                party = models.Party(name=name, role='author', email=email, url=url)
-                parties.append(party)
-            elif isinstance(author, str):
-                parties.append(models.Party(name=author, role='author'))
-            else:
-                parties.append(models.Party(name=repr(author), role='author'))
-
-        homepage_url = package_data.get('homepage')
-
-        repository = package_data.get('repository') or {}
-        repo_type = repository.get('type')
-        repo_url = repository.get('url')
-
-        vcs_url = None
-        if repo_type and repo_url:
-            vcs_url = f'{repo_type}+{repo_url}'
 
         deps = package_data.get('dependencies') or {}
         dependencies = []
@@ -86,17 +57,52 @@ def parse(cls, location):
                     is_optional=True,
                 )
             )
-
-        yield models.PackageData(
+        
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
             name=name,
-            description=description,
             version=version,
-            extracted_license_statement=extracted_license_statement,
-            keywords=keywords,
-            parties=parties,
-            homepage_url=homepage_url,
-            vcs_url=vcs_url,
-            dependencies=dependencies
+            dependencies=dependencies,
         )
+        if purl_only:
+            yield pkg
+            return
+
+        description = package_data.get('description')
+        extracted_license_statement = package_data.get('license')
+        keywords = package_data.get('keywords') or []
+
+        parties = []
+
+        authors = package_data.get('authors') or []
+        for author in authors:
+            if isinstance(author, dict):
+                name = author.get('name')
+                email = author.get('email')
+                url = author.get('homepage')
+                party = models.Party(name=name, role='author', email=email, url=url)
+                parties.append(party)
+            elif isinstance(author, str):
+                parties.append(models.Party(name=author, role='author'))
+            else:
+                parties.append(models.Party(name=repr(author), role='author'))
+
+        homepage_url = package_data.get('homepage')
+
+        repository = package_data.get('repository') or {}
+        repo_type = repository.get('type')
+        repo_url = repository.get('url')
+
+        vcs_url = None
+        if repo_type and repo_url:
+            vcs_url = f'{repo_type}+{repo_url}'
+
+        pkg.description = description
+        pkg.primary_language = BowerJsonHandler.default_primary_language
+        pkg.extracted_license_statement = extracted_license_statement
+        pkg.keywords = keywords
+        pkg.parties = parties
+        pkg.homepage_url = homepage_url
+        pkg.vcs_url = vcs_url
+        yield pkg
diff --git a/src/packagedcode/build.py b/src/packagedcode/build.py
index 7657f7c64b3..768fea3ea93 100644
--- a/src/packagedcode/build.py
+++ b/src/packagedcode/build.py
@@ -55,7 +55,7 @@ class AutotoolsConfigureHandler(models.DatafileHandler):
     documentation_url = 'https://www.gnu.org/software/automake/'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, **kwargs):
         # we use the parent directory as a package name
         name = fileutils.file_name(fileutils.parent_directory(location))
         # we could use checksums as version in the future
@@ -143,7 +143,7 @@ def assemble(cls, package_data, resource, codebase, package_adder):
         yield resource
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
 
         # Thanks to Starlark being a Python dialect, we can use `ast` to parse it
         with open(location, 'rb') as f:
@@ -191,18 +191,20 @@ def parse(cls, location):
                     if not name:
                         continue
 
-                    license_files = args.get('licenses')
-
-                    if TRACE:
-                        logger_debug(f"build: parse: license_files: {license_files}")
-
                     package_data = models.PackageData(
                         datasource_id=cls.datasource_id,
                         type=cls.default_package_type,
                         name=name,
                     )
 
-                    package_data.extracted_license_statement = license_files
+                    if not purl_only:
+                        license_files = args.get('licenses')
+                        package_data.extracted_license_statement = license_files
+                        if TRACE:
+                            logger_debug(
+                                f"build: parse: license_files: {license_files}"
+                            )
+
                     yield package_data
 
         else:
@@ -334,7 +336,7 @@ class BuckMetadataBzlHandler(BaseStarlarkManifestHandler):
     documentation_url = 'https://buck.build/'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
 
         with open(location, 'rb') as f:
             tree = ast.parse(f.read())
@@ -366,15 +368,16 @@ def parse(cls, location):
                     metadata_fields[key_name] = value
 
         parties = []
-        maintainers = metadata_fields.get('maintainers', []) or []
-        for maintainer in maintainers:
-            parties.append(
-                models.Party(
-                    type=models.party_org,
-                    name=maintainer,
-                    role='maintainer',
+        if not purl_only:
+            maintainers = metadata_fields.get('maintainers', []) or []
+            for maintainer in maintainers:
+                parties.append(
+                    models.Party(
+                        type=models.party_org,
+                        name=maintainer,
+                        role='maintainer',
+                    )
                 )
-            )
 
         if (
             'upstream_type'
@@ -386,16 +389,19 @@ def parse(cls, location):
         ):
             # TODO: Create function that determines package type from download URL,
             # then create a package of that package type from the metadata info
-            yield models.PackageData(
+            
+            pkg = models.PackageData(
                 datasource_id=cls.datasource_id,
                 type=metadata_fields.get('upstream_type', cls.default_package_type),
                 name=metadata_fields.get('name'),
                 version=metadata_fields.get('version'),
-                extracted_license_statement=metadata_fields.get('licenses', []),
-                parties=parties,
-                homepage_url=metadata_fields.get('upstream_address', ''),
                 # TODO: Store 'upstream_hash` somewhere
             )
+            if not purl_only:
+                pkg.extracted_license_statement = metadata_fields.get('licenses', [])
+                pkg.parties = parties
+                pkg.homepage_url = metadata_fields.get('upstream_address', '')
+            yield pkg
 
         if (
             'package_type'
@@ -409,19 +415,20 @@ def parse(cls, location):
             and 'vcs_commit_hash'
             in metadata_fields
         ):
-            yield models.PackageData(
+            pkg = models.PackageData(
                 datasource_id=cls.datasource_id,
                 type=metadata_fields.get('package_type', cls.default_package_type),
                 name=metadata_fields.get('name'),
                 version=metadata_fields.get('version'),
-                extracted_license_statement=metadata_fields.get('license_expression', ''),
-                parties=parties,
-                homepage_url=metadata_fields.get('homepage_url', ''),
-                download_url=metadata_fields.get('download_url', ''),
-                vcs_url=metadata_fields.get('vcs_url', ''),
-                sha1=metadata_fields.get('download_archive_sha1', ''),
-                extra_data=dict(vcs_commit_hash=metadata_fields.get('vcs_commit_hash', ''))
             )
+            pkg.extracted_license_statement = metadata_fields.get('license_expression', '')
+            pkg.parties = parties
+            pkg.homepage_url = metadata_fields.get('homepage_url', '')
+            pkg.download_url = metadata_fields.get('download_url', '')
+            pkg.vcs_url = metadata_fields.get('vcs_url', '')
+            pkg.sha1 = metadata_fields.get('download_archive_sha1', '')
+            pkg.extra_data = dict(vcs_commit_hash=metadata_fields.get('vcs_commit_hash', ''))
+            yield pkg
 
     @classmethod
     def assign_package_to_resources(cls, package, resource, codebase, package_adder):
diff --git a/src/packagedcode/build_gradle.py b/src/packagedcode/build_gradle.py
index 308abe5bc62..51669a4b962 100644
--- a/src/packagedcode/build_gradle.py
+++ b/src/packagedcode/build_gradle.py
@@ -59,9 +59,9 @@ class BuildGradleHandler(models.DatafileHandler):
     description = 'Gradle build script'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         dependencies = get_dependencies(location)
-        return build_package(cls, dependencies)
+        return build_package(cls, dependencies, purl_only=purl_only)
 
     # TODO: handle complex cases of nested builds with many packages
     @classmethod
@@ -328,7 +328,7 @@ def get_dependencies(build_gradle_location):
     return list(get_dependencies_from_parse_tree(parse_tree))
 
 
-def build_package(cls, dependencies):
+def build_package(cls, dependencies, purl_only=False):
     """
     Yield PackageData from a ``dependencies`` list of mappings.
     """
@@ -364,10 +364,12 @@ def build_package(cls, dependencies):
             )
         )
 
-    yield models.PackageData(
+
+    pkg = models.PackageData(
         datasource_id=cls.datasource_id,
         type=cls.default_package_type,
-        primary_language=BuildGradleHandler.default_primary_language,
         dependencies=package_dependencies,
     )
-
+    if purl_only:
+        pkg.primary_language = BuildGradleHandler.default_primary_language
+    yield pkg
\ No newline at end of file
diff --git a/src/packagedcode/cargo.py b/src/packagedcode/cargo.py
index ffe8dea5159..a6ddbd9a382 100644
--- a/src/packagedcode/cargo.py
+++ b/src/packagedcode/cargo.py
@@ -29,13 +29,31 @@ class CargoTomlHandler(models.DatafileHandler):
     documentation_url = 'https://doc.rust-lang.org/cargo/reference/manifest.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         package_data = toml.load(location, _dict=dict)
 
         core_package_data = package_data.get('package', {})
 
         name = core_package_data.get('name')
         version = core_package_data.get('version')
+
+        # cargo dependencies are complex and can be overriden at multiple levels
+        dependencies = []
+        for key, value in core_package_data.items():
+            if key.endswith('dependencies'):
+                dependencies.extend(dependency_mapper(dependencies=value, scope=key))
+
+        pkg = models.PackageData(
+            datasource_id=cls.datasource_id,
+            type=cls.default_package_type,
+            name=name,
+            version=version,
+            dependencies=dependencies,
+        )
+        if purl_only:
+            yield pkg
+            return
+
         description = core_package_data.get('description') or ''
         description = description.strip()
 
@@ -50,12 +68,6 @@ def parse(cls, location):
         categories = core_package_data.get('categories') or []
         keywords.extend(categories)
 
-        # cargo dependencies are complex and can be overriden at multiple levels
-        dependencies = []
-        for key, value in core_package_data.items():
-            if key.endswith('dependencies'):
-                dependencies.extend(dependency_mapper(dependencies=value, scope=key))
-
         # TODO: add file refs:
         # - readme, include and exclude
         # TODO: other URLs
@@ -67,22 +79,17 @@ def parse(cls, location):
         repository_download_url = name and version and f'https://crates.io/api/v1/crates/{name}/{version}/download'
         api_data_url = name and f'https://crates.io/api/v1/crates/{name}'
 
-        yield models.PackageData(
-            datasource_id=cls.datasource_id,
-            type=cls.default_package_type,
-            name=name,
-            version=version,
-            primary_language=cls.default_primary_language,
-            description=description,
-            parties=parties,
-            extracted_license_statement=extracted_license_statement,
-            vcs_url=vcs_url,
-            homepage_url=homepage_url,
-            repository_homepage_url=repository_homepage_url,
-            repository_download_url=repository_download_url,
-            api_data_url=api_data_url,
-            dependencies=dependencies,
-        )
+        pkg.primary_language = cls.default_primary_language
+        pkg.description = description
+        pkg.parties = parties
+        pkg.extracted_license_statement = extracted_license_statement
+        pkg.vcs_url = vcs_url
+        pkg.homepage_url = homepage_url
+        pkg.repository_homepage_url = repository_homepage_url
+        pkg.repository_download_url = repository_download_url
+        pkg.api_data_url = api_data_url
+        pkg.dependencies = dependencies
+        yield pkg
 
     @classmethod
     def assemble(cls, package_data, resource, codebase, package_adder):
@@ -116,7 +123,7 @@ class CargoLockHandler(models.DatafileHandler):
     # ]
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         cargo_lock = toml.load(location, _dict=dict)
         dependencies = []
         package = cargo_lock.get('package', [])
@@ -137,12 +144,14 @@ def parse(cls, location):
                 )
             )
 
-        yield models.PackageData(
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
             dependencies=dependencies,
         )
+        if not purl_only:
+            pkg.primary_language = cls.default_primary_language
+        yield pkg
 
     @classmethod
     def assemble(cls, package_data, resource, codebase, package_adder):
diff --git a/src/packagedcode/chef.py b/src/packagedcode/chef.py
index 5c378cdce4e..b5319ddd341 100644
--- a/src/packagedcode/chef.py
+++ b/src/packagedcode/chef.py
@@ -183,14 +183,18 @@ def is_datafile(cls, location, filetypes=tuple()):
             return not parent.endswith('dist-info')
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         """
         Yield one or more Package manifest objects given a file ``location``
         pointing to a package archive, manifest or similar.
         """
         with io.open(location, encoding='utf-8') as loc:
             package_data = json.load(loc)
-        yield build_package(package_data, datasource_id=cls.datasource_id)
+        yield build_package(
+            package_data,
+            datasource_id=cls.datasource_id,
+            purl_only=purl_only,
+        )
 
 
 class ChefMetadataRbHandler(BaseChefMetadataHandler):
@@ -202,7 +206,7 @@ class ChefMetadataRbHandler(BaseChefMetadataHandler):
     documentation_url = 'https://docs.chef.io/config_rb_metadata/'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with io.open(location, encoding='utf-8') as loc:
             file_contents = loc.read()
 
@@ -213,10 +217,14 @@ def parse(cls, location):
             ChefMetadataFormatter()
         )
         package_data = json.loads(formatted_file_contents)
-        yield build_package(package_data, datasource_id=cls.datasource_id)
+        yield build_package(
+            package_data,
+            datasource_id=cls.datasource_id,
+            purl_only=purl_only,
+        )
 
 
-def build_package(package_data, datasource_id):
+def build_package(package_data, datasource_id, purl_only=False):
     """
     Return a PackageData object from a package_data mapping from a metadata.json
     or similar or None.
@@ -224,6 +232,32 @@ def build_package(package_data, datasource_id):
     name = package_data.get('name')
     version = package_data.get('version')
 
+    deps = dict(package_data.get('dependencies', {}) or {})
+    deps.update(package_data.get('depends', {}) or {})
+
+    dependencies = []
+    for dependency_name, requirement in deps.items():
+        dependencies.append(
+            models.DependentPackage(
+                purl=PackageURL(type='chef', name=dependency_name).to_string(),
+                scope='dependencies',
+                extracted_requirement=requirement,
+                is_runtime=True,
+                is_optional=False,
+            )
+        )
+    
+    pkg = models.PackageData(
+        datasource_id=datasource_id,
+        type=ChefMetadataJsonHandler.default_package_type,
+        name=name,
+        version=version,
+        dependencies=dependencies,
+    )
+    if purl_only:
+        yield pkg
+        return
+
     maintainer_name = package_data.get('maintainer', '')
     maintainer_email = package_data.get('maintainer_email', '')
     parties = []
@@ -246,32 +280,16 @@ def build_package(package_data, datasource_id):
     code_view_url = package_data.get('source_url', '')
     bug_tracking_url = package_data.get('issues_url', '')
 
-    deps = dict(package_data.get('dependencies', {}) or {})
-    deps.update(package_data.get('depends', {}) or {})
-
-    dependencies = []
-    for dependency_name, requirement in deps.items():
-        dependencies.append(
-            models.DependentPackage(
-                purl=PackageURL(type='chef', name=dependency_name).to_string(),
-                scope='dependencies',
-                extracted_requirement=requirement,
-                is_runtime=True,
-                is_optional=False,
-            )
-        )
-
-    return models.PackageData(
-        datasource_id=datasource_id,
-        type=ChefMetadataJsonHandler.default_package_type,
-        name=name,
-        version=version,
-        parties=parties,
-        description=description.strip() or None,
-        extracted_license_statement=extracted_license_statement,
-        code_view_url=code_view_url.strip() or None,
-        bug_tracking_url=bug_tracking_url.strip() or None,
-        dependencies=dependencies,
-        primary_language='Ruby',
-        **get_urls(name, version),
-    )
+    pkg.parties = parties
+    pkg.description = description.strip() or None
+    pkg.extracted_license_statement = extracted_license_statement
+    pkg.code_view_url = code_view_url.strip() or None
+    pkg.bug_tracking_url = bug_tracking_url.strip() or None
+    pkg.primary_language = 'Ruby'
+    (
+        pkg.download_url,
+        pkg.repository_download_url,
+        pkg.repository_homepage_url,
+        pkg.api_data_url
+    ) = get_urls(name, version)
+    yield pkg
diff --git a/src/packagedcode/cocoapods.py b/src/packagedcode/cocoapods.py
index 9773ae13638..a0a9c3fd8c0 100644
--- a/src/packagedcode/cocoapods.py
+++ b/src/packagedcode/cocoapods.py
@@ -216,7 +216,7 @@ class PodspecHandler(BasePodHandler):
     documentation_url = 'https://guides.cocoapods.org/syntax/podspec.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         """
         Yield one or more Package manifest objects given a file ``location``
         pointing to a package archive, manifest or similar.
@@ -224,10 +224,24 @@ def parse(cls, location):
         podspec = spec.parse_spec(
             location=location,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
         name = podspec.get('name')
         version = podspec.get('version')
+        dependencies = podspec.get('dependencies')
+
+        pkg = models.PackageData(
+            datasource_id=cls.datasource_id,
+            type=cls.default_package_type,
+            name=name,
+            version=version,
+            dependencies=dependencies,
+        )
+        if purl_only:
+            yield pkg
+            return
+
         homepage_url = podspec.get('homepage')
         extracted_license_statement = podspec.get('license')
         summary = podspec.get('summary')
@@ -252,27 +266,27 @@ def parse(cls, location):
                 )
                 parties.append(party)
 
-        urls = get_urls(
+        (
+            pkg.repository_download_url,
+            pkg.repository_homepage_url,
+            pkg.code_view_url,
+            pkg.bug_tracking_url,
+            pkg.api_data_url,
+        ) = get_urls(
             name=name,
             version=version,
             homepage_url=homepage_url,
-            vcs_url=vcs_url)
-
-        yield models.PackageData(
-            datasource_id=cls.datasource_id,
-            type=cls.default_package_type,
-            name=name,
-            version=version,
-            primary_language=cls.default_primary_language,
             vcs_url=vcs_url,
-            # FIXME: a source should be a PURL, not a list of URLs
-            # source_packages=vcs_url.split('\n'),
-            description=description,
-            extracted_license_statement=extracted_license_statement,
-            homepage_url=homepage_url,
-            parties=parties,
-            **urls,
         )
+        pkg.primary_language = cls.default_primary_language
+        pkg.vcs_url = vcs_url
+        # FIXME: a source should be a PURL, not a list of URLs
+        # source_packages=vcs_url.split('\n'),
+        pkg.description = description
+        pkg.extracted_license_statement = extracted_license_statement
+        pkg.homepage_url = homepage_url
+        pkg.parties = parties
+        yield pkg
 
 
 class PodfileHandler(PodspecHandler):
@@ -293,7 +307,7 @@ class PodfileLockHandler(BasePodHandler):
     documentation_url = 'https://guides.cocoapods.org/using/the-podfile.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         """
         Yield PackageData from a YAML Podfile.lock.
         """
@@ -337,12 +351,14 @@ def parse(cls, location):
                     )
                 )
 
-        yield models.PackageData(
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
             dependencies=dependencies,
         )
+        if not purl_only:
+            pkg.primary_language = cls.default_primary_language
+        yield pkg
 
 
 class PodspecJsonHandler(models.DatafileHandler):
@@ -354,12 +370,27 @@ class PodspecJsonHandler(models.DatafileHandler):
     documentation_url = 'https://guides.cocoapods.org/syntax/podspec.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with open(location) as psj:
             data = json.load(psj)
 
         name = data.get('name')
         version = data.get('version')
+        dependencies = data.get('dependencies', '')
+        if dependencies:
+            extra_data['dependencies'] = dependencies
+
+        pkg = models.PackageData(
+            datasource_id=cls.datasource_id,
+            primary_language=cls.default_primary_language,
+            type=cls.default_package_type,
+            name=name,
+            version=version,
+        )
+        if purl_only:
+            yield pkg
+            return
+
         summary = data.get('summary', '')
         description = data.get('description', '')
         homepage_url = data.get('homepage')
@@ -414,29 +445,25 @@ def parse(cls, location):
 
         extra_data = {}
         extra_data['source'] = data['source']
-        dependencies = data.get('dependencies', '')
-        if dependencies:
-            extra_data['dependencies'] = dependencies
         extra_data['podspec.json'] = data
 
-        urls = get_urls(
-            name=name,
-            version=version, homepage_url=homepage_url, vcs_url=vcs_url)
-
-        yield models.PackageData(
-            datasource_id=cls.datasource_id,
-            primary_language=cls.default_primary_language,
-            type=cls.default_package_type,
+        (
+            pkg.repository_download_url,
+            pkg.repository_homepage_url,
+            pkg.code_view_url,
+            pkg.bug_tracking_url,
+            pkg.api_data_url,
+        ) = get_urls(
             name=name,
             version=version,
-            description=description,
-            extracted_license_statement=extracted_license_statement,
-            parties=parties,
-            vcs_url=vcs_url,
             homepage_url=homepage_url,
-            download_url=download_url,
-            **urls,
+            vcs_url=vcs_url,
         )
+        pkg.primary_language = cls.default_primary_language
+        pkg.description = description
+        pkg.extracted_license_statement = extracted_license_statement
+        pkg.parties = parties
+        yield pkg
 
 
 def get_urls(name=None, version=None, homepage_url=None, vcs_url=None, **kwargs):
diff --git a/src/packagedcode/conda.py b/src/packagedcode/conda.py
index 545b3a1fb53..1715e20121c 100644
--- a/src/packagedcode/conda.py
+++ b/src/packagedcode/conda.py
@@ -79,7 +79,7 @@ def assign_package_to_resources(cls, package, resource, codebase, package_adder)
         )
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         metayaml = get_meta_yaml_data(location)
         package_element = metayaml.get('package') or {}
         package_name = package_element.get('name')
@@ -87,17 +87,6 @@ def parse(cls, location):
             return
         version = package_element.get('version')
 
-        # FIXME: source is source, not download
-        source = metayaml.get('source') or {}
-        download_url = source.get('url')
-        sha256 = source.get('sha256')
-
-        about = metayaml.get('about') or {}
-        homepage_url = about.get('home')
-        extracted_license_statement = about.get('license')
-        description = about.get('summary')
-        vcs_url = about.get('dev_url')
-
         dependencies = []
         requirements = metayaml.get('requirements') or {}
         for scope, reqs in requirements.items():
@@ -118,20 +107,36 @@ def parse(cls, location):
                     )
                 )
 
-        yield models.PackageData(
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
             name=package_name,
             version=version,
-            download_url=download_url,
-            homepage_url=homepage_url,
-            vcs_url=vcs_url,
-            description=description,
-            sha256=sha256,
-            extracted_license_statement=extracted_license_statement,
             dependencies=dependencies,
         )
 
+        if purl_only:
+            yield pkg
+            return
+
+        # FIXME: source is source, not download
+        source = metayaml.get('source') or {}
+        download_url = source.get('url')
+        sha256 = source.get('sha256')
+
+        about = metayaml.get('about') or {}
+        homepage_url = about.get('home')
+        extracted_license_statement = about.get('license')
+        description = about.get('summary')
+        vcs_url = about.get('dev_url')
+
+        pkg.download_url = download_url
+        pkg.homepage_url = homepage_url
+        pkg.vcs_url = vcs_url
+        pkg.description = description
+        pkg.sha256 = sha256
+        pkg.extracted_license_statement = extracted_license_statement
+        yield pkg
 
 def get_meta_yaml_data(location):
     """
diff --git a/src/packagedcode/cran.py b/src/packagedcode/cran.py
index ef274105da6..ea68af26467 100644
--- a/src/packagedcode/cran.py
+++ b/src/packagedcode/cran.py
@@ -30,38 +30,12 @@ class CranDescriptionFileHandler(models.DatafileHandler):
     documentation_url = 'https://r-pkgs.org/description.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         cran_desc = get_cran_description(location)
-
         name = cran_desc.get('Package')
         if not name:
             return
 
-        parties = []
-        maintainers = cran_desc.get('Maintainer') or ''
-        for maintainer in maintainers.split(',\n'):
-            maintainer_name, maintainer_email = get_party_info(maintainer)
-            if maintainer_name or maintainer_email:
-                parties.append(
-                    models.Party(
-                        name=maintainer_name,
-                        role='maintainer',
-                        email=maintainer_email,
-                    )
-                )
-
-        authors = cran_desc.get('Author') or ''
-        for author in authors.split(',\n'):
-            author_name, author_email = get_party_info(author)
-            if author_name or author_email:
-                parties.append(
-                    models.Party(
-                        name=author_name,
-                        role='author',
-                        email=author_email,
-                    )
-                )
-
         package_dependencies = []
         dependencies = cran_desc.get('Depends') or ''
         for dependency in dependencies.split(',\n'):
@@ -88,23 +62,53 @@ def parse(cls, location):
                 )
             )
 
-        extracted_license_statement = cran_desc.get('License')
-
-        # TODO: Let's handle the release date as a Date type
-        # release_date = cran_desc.get('Date/Publication'),
-
-        yield models.PackageData(
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
             name=name,
             version=cran_desc.get('Version'),
-            # TODO: combine both together
-            description=cran_desc.get('Description', '') or cran_desc.get('Title', ''),
-            extracted_license_statement=extracted_license_statement,
-            parties=parties,
             dependencies=package_dependencies,
-            repository_homepage_url=f'https://cran.r-project.org/package={name}',
         )
+        if purl_only:
+            yield pkg
+            return
+
+        parties = []
+        maintainers = cran_desc.get('Maintainer') or ''
+        for maintainer in maintainers.split(',\n'):
+            maintainer_name, maintainer_email = get_party_info(maintainer)
+            if maintainer_name or maintainer_email:
+                parties.append(
+                    models.Party(
+                        name=maintainer_name,
+                        role='maintainer',
+                        email=maintainer_email,
+                    )
+                )
+
+        authors = cran_desc.get('Author') or ''
+        for author in authors.split(',\n'):
+            author_name, author_email = get_party_info(author)
+            if author_name or author_email:
+                parties.append(
+                    models.Party(
+                        name=author_name,
+                        role='author',
+                        email=author_email,
+                    )
+                )
+
+        extracted_license_statement = cran_desc.get('License')
+
+        # TODO: Let's handle the release date as a Date type
+        # release_date = cran_desc.get('Date/Publication'),
+
+        # TODO: combine both together
+        pkg.description = cran_desc.get('Description', '') or cran_desc.get('Title', '')
+        pkg.extracted_license_statement = extracted_license_statement
+        pkg.parties = parties
+        pkg.repository_homepage_url = f'https://cran.r-project.org/package={name}'
+        yield pkg
 
 # FIXME: THIS IS NOT YAML but RFC 822
 
diff --git a/src/packagedcode/debian.py b/src/packagedcode/debian.py
index f84fc9e9afe..37321454069 100644
--- a/src/packagedcode/debian.py
+++ b/src/packagedcode/debian.py
@@ -59,11 +59,12 @@ class DebianDebPackageHandler(models.DatafileHandler):
     documentation_url = 'https://manpages.debian.org/unstable/dpkg-dev/deb.5.en.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         yield build_package_data_from_package_filename(
             filename=fileutils.file_name(location),
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -82,13 +83,14 @@ class DebianSourcePackageMetadataTarballHandler(models.DatafileHandler):
     documentation_url = 'https://manpages.debian.org/unstable/dpkg-dev/deb.5.en.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         # strip extension
         filename, _, _ = location.rpartition('.tar')
         yield build_package_data_from_package_filename(
             filename=filename,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -107,13 +109,14 @@ class DebianSourcePackageTarballHandler(models.DatafileHandler):
     documentation_url = 'https://manpages.debian.org/unstable/dpkg-dev/deb.5.en.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         # strip extension
         filename, _, _ = location.rpartition('.tar')
         yield build_package_data_from_package_filename(
             filename=filename,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -131,12 +134,13 @@ class DebianControlFileInExtractedDebHandler(models.DatafileHandler):
     documentation_url = 'https://www.debian.org/doc/debian-policy/ch-controlfields.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         # TODO: we cannot know the distro from the name only
         yield build_package_data(
             debian_data=get_paragraph_data_from_file(location=location),
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -156,7 +160,7 @@ class DebianControlFileInSourceHandler(models.DatafileHandler):
     documentation_url = 'https://www.debian.org/doc/debian-policy/ch-controlfields.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         # TODO: we cannot know the distro from the name only
         # NOTE: a control file in a source repo or debina.tar tarball can contain more than one package
         for debian_data in get_paragraphs_data_from_file(location=location):
@@ -164,6 +168,7 @@ def parse(cls, location):
                 debian_data,
                 datasource_id=cls.datasource_id,
                 package_type=cls.default_package_type,
+                purl_only=purl_only,
             )
 
     @classmethod
@@ -185,7 +190,7 @@ class DebianDscFileHandler(models.DatafileHandler):
     documentation_url = 'https://wiki.debian.org/dsc'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         # this is typically signed
         debian_data = get_paragraph_data_from_file(
             location=location,
@@ -195,6 +200,7 @@ def parse(cls, location):
             debian_data=debian_data,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -211,7 +217,7 @@ class DebianInstalledStatusDatabaseHandler(models.DatafileHandler):
     documentation_url = 'https://www.debian.org/doc/debian-policy/ch-controlfields.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         # note that we do not know yet the distro at this stage
         # we could get it... but we get that later during assemble()
         for debian_data in get_paragraphs_data_from_file(location):
@@ -219,6 +225,7 @@ def parse(cls, location):
                 debian_data,
                 datasource_id=cls.datasource_id,
                 package_type=cls.default_package_type,
+                purl_only=purl_only,
             )
 
     @classmethod
@@ -372,7 +379,7 @@ class DebianDistrolessInstalledDatabaseHandler(models.DatafileHandler):
     documentation_url = 'https://www.debian.org/doc/debian-policy/ch-controlfields.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         """
         Yield installed PackageData objects given a ``location``
         var/lib/dpkg/status.d/<status> file as found in a distroless container
@@ -385,6 +392,7 @@ def parse(cls, location):
                 datasource_id=cls.datasource_id,
                 package_type=cls.default_package_type,
                 distro='distroless',
+                purl_only=purl_only,
             )
 
     @classmethod
@@ -451,11 +459,12 @@ class DebianInstalledFilelistHandler(models.DatafileHandler):
     description = 'Debian installed file paths list'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         return parse_debian_files_list(
             location=location,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -477,11 +486,12 @@ class DebianInstalledMd5sumFilelistHandler(models.DatafileHandler):
     documentation_url = 'https://www.debian.org/doc/manuals/debian-handbook/sect.package-meta-information.en.html#sect.configuration-scripts'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         return parse_debian_files_list(
             location=location,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -502,11 +512,12 @@ class DebianMd5sumFilelistInPackageHandler(models.DatafileHandler):
     documentation_url = 'https://www.debian.org/doc/manuals/debian-handbook/sect.package-meta-information.en.html#sect.configuration-scripts'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         return parse_debian_files_list(
             location=location,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -517,7 +528,12 @@ def assign_package_to_resources(cls, package, resource, codebase, package_adder)
             return models.DatafileHandler.assign_package_to_resources(package, root, codebase, package_adder)
 
 
-def build_package_data_from_package_filename(filename, datasource_id, package_type):
+def build_package_data_from_package_filename(
+    filename,
+    datasource_id,
+    package_type,
+    **kwargs,
+):
     """
     Return a PackageData built from the filename of a Debian package archive.
     """
@@ -543,7 +559,12 @@ def build_package_data_from_package_filename(filename, datasource_id, package_ty
     )
 
 
-def parse_debian_files_list(location, datasource_id, package_type):
+def parse_debian_files_list(
+    location,
+    datasource_id,
+    package_type,
+    purl_only=False,
+):
     """
     Yield PackageData from a list of file paths at locations such as an from a
     Debian installed .list or .md5sums file.
@@ -578,16 +599,26 @@ def parse_debian_files_list(location, datasource_id, package_type):
     if not file_references:
         return
 
-    yield models.PackageData(
+    package = models.PackageData(
         datasource_id=datasource_id,
         type=package_type,
         name=name,
         qualifiers=qualifiers,
-        file_references=file_references,
     )
+    if purl_only:
+        yield package
+    else:
+        package.file_references = file_references
+        yield package
 
 
-def build_package_data(debian_data, datasource_id, package_type='deb', distro=None):
+def build_package_data(
+    debian_data,
+    datasource_id,
+    package_type='deb',
+    distro=None,
+    purl_only=False,
+):
     """
     Return a PackageData object from a package_data mapping (from a dpkg status
     or similar file) or None.
@@ -600,6 +631,17 @@ def build_package_data(debian_data, datasource_id, package_type='deb', distro=No
     if architecture:
         qualifiers['architecture'] = architecture
 
+    package = models.PackageData(
+        datasource_id=datasource_id,
+        type=package_type,
+        namespace=distro,
+        name=name,
+        version=version,
+        qualifiers=qualifiers,
+    )
+    if purl_only:
+        return package
+
     extra_data = {}
     # Multi-Arch can be: "foreign", "same", "allowed", "all", "optional" or
     # empty/non-present. See https://wiki.debian.org/Multiarch/HOWTO
@@ -639,21 +681,14 @@ def build_package_data(debian_data, datasource_id, package_type='deb', distro=No
 
         source_packages.append(source_pkg_purl)
 
-    return models.PackageData(
-        datasource_id=datasource_id,
-        type=package_type,
-        namespace=distro,
-        name=name,
-        version=version,
-        qualifiers=qualifiers,
-        description=description,
-        homepage_url=homepage_url,
-        size=size,
-        source_packages=source_packages,
-        keywords=keywords,
-        parties=parties,
-        extra_data=extra_data,
-    )
+    package.description = description
+    package.homepage_url = homepage_url
+    package.size = size
+    package.source_packages = source_packages
+    package.keywords = keywords
+    package.parties = parties
+    package.extra_data = extra_data
+    return package
 
 
 ignored_root_dirs = {
diff --git a/src/packagedcode/debian_copyright.py b/src/packagedcode/debian_copyright.py
index fe7aebc852d..f06a8246eee 100644
--- a/src/packagedcode/debian_copyright.py
+++ b/src/packagedcode/debian_copyright.py
@@ -93,12 +93,8 @@ def is_datafile(cls, location, filetypes=tuple(), strict=False):
                 return True
 
     @classmethod
-    def parse(cls, location):
-        debian_copyright = parse_copyright_file(location)
-        license_fields = DebianLicenseFields.get_license_fields(
-            debian_copyright=debian_copyright
-        )
-
+    def parse(cls, location, purl_only=False):
+        
         # TODO: collect the upstream source package details
 
         # find a name... TODO: this should be pushed down to each handler
@@ -109,19 +105,29 @@ def parse(cls, location):
             # no name otherwise for now
             name = None
 
-        yield models.PackageData(
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
             name=name,
-            extracted_license_statement=license_fields.extracted_license_statement,
-            declared_license_expression=license_fields.declared_license_expression,
-            declared_license_expression_spdx=license_fields.declared_license_expression_spdx,
-            license_detections=license_fields.license_detections,
-            other_license_expression=license_fields.other_license_expression,
-            other_license_expression_spdx=license_fields.other_license_expression_spdx,
-            other_license_detections=license_fields.other_license_detections,
-            copyright=debian_copyright.get_copyright(),
         )
+        if purl_only:
+            yield pkg
+            return
+
+        debian_copyright = parse_copyright_file(location)
+        license_fields = DebianLicenseFields.get_license_fields(
+            debian_copyright=debian_copyright
+        )
+
+        pkg.extracted_license_statement = license_fields.extracted_license_statement
+        pkg.declared_license_expression = license_fields.declared_license_expression
+        pkg.declared_license_expression_spdx = license_fields.declared_license_expression_spdx
+        pkg.license_detections = license_fields.license_detections
+        pkg.other_license_expression = license_fields.other_license_expression
+        pkg.other_license_expression_spdx = license_fields.other_license_expression_spdx
+        pkg.other_license_detections = license_fields.other_license_detections
+        pkg.copyright = debian_copyright.get_copyright()
+        yield pkg
 
 
 @attr.s
diff --git a/src/packagedcode/distro.py b/src/packagedcode/distro.py
index 6336e6b3111..9c22dc5bf1d 100644
--- a/src/packagedcode/distro.py
+++ b/src/packagedcode/distro.py
@@ -25,7 +25,7 @@ class EtcOsReleaseHandler(models.NonAssemblableDatafileHandler):
     documentation_url = 'https://www.freedesktop.org/software/systemd/man/os-release.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, **kwargs):
         distro = Distro.from_os_release_file(location)
         distro_identifier = distro.identifier
         pretty_name = distro.pretty_name and distro.pretty_name.lower() or ''
diff --git a/src/packagedcode/freebsd.py b/src/packagedcode/freebsd.py
index c1f5342b562..885ef100a10 100644
--- a/src/packagedcode/freebsd.py
+++ b/src/packagedcode/freebsd.py
@@ -52,7 +52,7 @@ class CompactManifestHandler(models.DatafileHandler):
     documentation_url = 'https://www.freebsd.org/cgi/man.cgi?pkg-create(8)#MANIFEST_FILE_DETAILS'
 
     @classmethod
-    def _parse(cls, yaml_data):
+    def _parse(cls, yaml_data, purl_only=False):
         package_data = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
@@ -63,9 +63,23 @@ def _parse(cls, yaml_data):
         )
 
         # mapping of top level manifest items to the PackageData object field name
-        plain_fields = [
+        purl_fields = [
             ('name', 'name'),
             ('version', 'version'),
+        ]
+
+        for source, target in purl_fields:
+            value = yaml_data.get(source)
+            if value:
+                if isinstance(value, str):
+                    value = value.strip()
+                if value:
+                    setattr(package_data, target, value)
+
+        if purl_only:
+            return package_data
+
+        plain_fields = [
             ('www', 'homepage_url'),
             ('desc', 'description'),
             ('categories', 'keywords'),
@@ -96,7 +110,6 @@ def _parse(cls, yaml_data):
 
         # license_mapper needs multiple fields
         license_mapper(yaml_data, package_data)
-
         cls.populate_license_fields(package_data)
 
         if TRACE:
@@ -107,15 +120,15 @@ def _parse(cls, yaml_data):
         return package_data
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         """
-        Yield one or more Package manifest objects given a file ``location`` pointing to a
-        package archive, manifest or similar.
+        Yield one or more Package manifest objects given a file ``location``
+        pointing to a package archive, manifest or similar.
         """
         with io.open(location, encoding='utf-8') as loc:
             yaml_data = saneyaml.load(loc)
 
-        yield cls._parse(yaml_data)
+        yield cls._parse(yaml_data=yaml_data, purl_only=purl_only)
 
     @staticmethod
     def get_license_detections_and_expression(package_data):
diff --git a/src/packagedcode/godeps.py b/src/packagedcode/godeps.py
index 68ff54787ff..4bebfb7e055 100644
--- a/src/packagedcode/godeps.py
+++ b/src/packagedcode/godeps.py
@@ -38,7 +38,7 @@ class GodepsHandler(models.NonAssemblableDatafileHandler):
     documentation_url = 'https://github.com/tools/godep'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         godeps = Godep(location)
 
         if godeps.import_path:
@@ -64,14 +64,20 @@ def parse(cls, location):
                 )
             )
 
-        yield models.PackageData(
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
             namespace=namespace,
             name=name,
-            primary_language=cls.default_primary_language,
             dependencies=dependencies,
-         )
+        )
+
+        if purl_only:
+            yield pkg
+            return
+
+        pkg.primary_language = cls.default_primary_language
+        yield pkg
 
     @classmethod
     def assign_package_to_resources(cls, package, resource, codebase, package_adder):
diff --git a/src/packagedcode/golang.py b/src/packagedcode/golang.py
index 6075c713e86..21374b45f66 100644
--- a/src/packagedcode/golang.py
+++ b/src/packagedcode/golang.py
@@ -49,7 +49,7 @@ class GoModHandler(BaseGoModuleHandler):
     documentation_url = 'https://go.dev/ref/mod'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         gomods = go_mod.parse_gomod(location)
 
         dependencies = []
@@ -82,6 +82,18 @@ def parse(cls, location):
         name = gomods.name
         namespace = gomods.namespace
 
+        pkg = models.PackageData(
+            datasource_id=cls.datasource_id,
+            type=cls.default_package_type,
+            name=name,
+            namespace=namespace,
+            dependencies=dependencies,
+        )
+
+        if purl_only:
+            yield pkg
+            return
+
         homepage_url = f'https://pkg.go.dev/{gomods.namespace}/{gomods.name}'
         vcs_url = f'https://{gomods.namespace}/{gomods.name}.git'
 
@@ -89,17 +101,11 @@ def parse(cls, location):
         if namespace and name:
             repository_homepage_url = f'https://pkg.go.dev/{namespace}/{name}'
 
-        yield models.PackageData(
-            datasource_id=cls.datasource_id,
-            type=cls.default_package_type,
-            name=name,
-            namespace=namespace,
-            vcs_url=vcs_url,
-            homepage_url=homepage_url,
-            repository_homepage_url=repository_homepage_url,
-            dependencies=dependencies,
-            primary_language=cls.default_primary_language,
-        )
+        pkg.vcs_url = vcs_url
+        pkg.homepage_url = homepage_url
+        pkg.repository_homepage_url = repository_homepage_url
+        pkg.primary_language = cls.default_primary_language
+        yield pkg
 
 
 class GoSumHandler(BaseGoModuleHandler):
@@ -111,7 +117,7 @@ class GoSumHandler(BaseGoModuleHandler):
     documentation_url = 'https://go.dev/ref/mod#go-sum-files'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         gosums = go_mod.parse_gosum(location)
         package_dependencies = []
         for gosum in gosums:
@@ -126,9 +132,14 @@ def parse(cls, location):
                 )
             )
 
-        yield models.PackageData(
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
             dependencies=package_dependencies,
-            primary_language=cls.default_primary_language,
         )
+
+        if purl_only:
+            yield pkg
+        else:
+            pkg.primary_language = cls.default_primary_language
+            yield pkg
diff --git a/src/packagedcode/haxe.py b/src/packagedcode/haxe.py
index 99f5be4ef7d..961792313de 100644
--- a/src/packagedcode/haxe.py
+++ b/src/packagedcode/haxe.py
@@ -45,7 +45,7 @@ class HaxelibJsonHandler(models.DatafileHandler):
     documentation_url = 'https://lib.haxe.org/documentation/creating-a-haxelib-package/'
 
     @classmethod
-    def _parse(cls, json_data):
+    def _parse(cls, json_data, purl_only=False):
         name = json_data.get('name')
         version = json_data.get('version')
 
@@ -54,13 +54,28 @@ def _parse(cls, json_data):
             type=cls.default_package_type,
             name=name,
             version=version,
-            homepage_url=json_data.get('url'),
-            extracted_license_statement=json_data.get('license'),
-            keywords=json_data.get('tags'),
-            description=json_data.get('description'),
-            primary_language=cls.default_primary_language,
         )
 
+        for dep_name, dep_version in json_data.get('dependencies', {}).items():
+            dep_version = dep_version and dep_version.strip()
+            is_resolved = bool(dep_version)
+            dep_purl = PackageURL(
+                type=cls.default_package_type,
+                name=dep_name,
+                version=dep_version
+            ).to_string()
+            dep = models.DependentPackage(purl=dep_purl, is_resolved=is_resolved,)
+            package_data.dependencies.append(dep)
+
+        if purl_only:
+            return package_data
+
+        package_data.homepage_url = json_data.get('url')
+        package_data.extracted_license_statement = json_data.get('license')
+        package_data.keywords = json_data.get('tags')
+        package_data.description = json_data.get('description')
+        package_data.primary_language = cls.default_primary_language
+
         if name and version:
             download_url = f'https://lib.haxe.org/p/{name}/{version}/download/'
             package_data.repository_download_url = download_url
@@ -77,21 +92,10 @@ def _parse(cls, json_data):
                 url='https://lib.haxe.org/u/{}'.format(contrib))
             package_data.parties.append(party)
 
-        for dep_name, dep_version in json_data.get('dependencies', {}).items():
-            dep_version = dep_version and dep_version.strip()
-            is_resolved = bool(dep_version)
-            dep_purl = PackageURL(
-                type=cls.default_package_type,
-                name=dep_name,
-                version=dep_version
-            ).to_string()
-            dep = models.DependentPackage(purl=dep_purl, is_resolved=is_resolved,)
-            package_data.dependencies.append(dep)
-
         return package_data
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         """
         Yield one or more Package manifest objects given a file ``location`` pointing to a
         package_data archive, manifest or similar.
@@ -111,4 +115,4 @@ def parse(cls, location):
         with io.open(location, encoding='utf-8') as loc:
             json_data = json.load(loc)
 
-        yield cls._parse(json_data)
+        yield cls._parse(json_data=json_data, purl_only=purl_only)
diff --git a/src/packagedcode/jar_manifest.py b/src/packagedcode/jar_manifest.py
index 48df5373eb1..8e56ce62ded 100644
--- a/src/packagedcode/jar_manifest.py
+++ b/src/packagedcode/jar_manifest.py
@@ -62,7 +62,7 @@ def parse_section(section):
     return data
 
 
-def get_normalized_java_manifest_data(manifest_mapping):
+def get_normalized_java_manifest_data(manifest_mapping, purl_only=False):
     """
     Return a mapping of package-like data normalized from a mapping of the
     `manifest_mapping` data mapping or None.
@@ -231,6 +231,10 @@ def dget(s):
     package['namespace'] = namespace
     package['name'] = name
     package['version'] = version
+
+    if purl_only:
+        return package
+
     package['description'] = description
 
     # licensing
diff --git a/src/packagedcode/maven.py b/src/packagedcode/maven.py
index b5d521dc44a..ab97bc70353 100644
--- a/src/packagedcode/maven.py
+++ b/src/packagedcode/maven.py
@@ -132,11 +132,14 @@ class JavaJarManifestHandler(MavenBasePackageHandler):
     documentation_url = 'https://docs.oracle.com/javase/tutorial/deployment/jar/manifestindex.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         sections = parse_manifest(location)
         if sections:
             main_section = sections[0]
-            manifest = get_normalized_java_manifest_data(main_section)
+            manifest = get_normalized_java_manifest_data(
+                manifest_mapping=main_section,
+                purl_only=purl_only
+            )
             if manifest:
                 yield models.PackageData(**manifest,)
 
@@ -206,13 +209,19 @@ def is_datafile(cls, location, filetypes=tuple()):
                     return True
 
     @classmethod
-    def parse(cls, location, base_url='https://repo1.maven.org/maven2'):
+    def parse(
+        cls,
+        location,
+        base_url='https://repo1.maven.org/maven2',
+        purl_only=False,
+    ):
         package_data = parse(
             location=location,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
             primary_language=cls.default_primary_language,
             base_url=base_url,
+            purl_only=purl_only,
         )
         if package_data:
             yield package_data
@@ -303,7 +312,7 @@ class MavenPomPropertiesHandler(models.NonAssemblableDatafileHandler):
     documentation_url = 'https://maven.apache.org/pom.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         """
         Yield PackageData from a pom.properties file (which is typically side-
         by-side with its pom file.)
@@ -313,27 +322,36 @@ def parse(cls, location):
             if TRACE:
                 logger.debug(f'MavenPomPropertiesHandler.parse: properties: {properties!r}')
             if properties:
-                yield from cls.parse_pom_properties(properties=properties) 
+                yield from cls.parse_pom_properties(
+                    properties=properties,
+                    purl_only=purl_only,
+                )
 
     @classmethod
-    def parse_pom_properties(cls, properties):
+    def parse_pom_properties(cls, properties, purl_only=False):
         namespace = properties.pop("groupId", None)
         name = properties.pop("artifactId", None)
         version = properties.pop("version", None)
-        if properties:
-            extra_data = dict(pom_properties=properties)
-        else:
-            extra_data = {}
 
-        yield models.PackageData(
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
             name=name,
             namespace=namespace,
             version=version,
-            extra_data=extra_data,
         )
+        if purl_only:
+            yield purl_only
+            return
+
+        if properties:
+            extra_data = dict(pom_properties=properties)
+        else:
+            extra_data = {}
+
+        pkg.primary_language = cls.default_primary_language
+        pkg.extra_data = extra_data
+        yield pkg
 
 
 def build_url(
@@ -1189,6 +1207,7 @@ def parse(
     package_type,
     primary_language,
     base_url='https://repo1.maven.org/maven2',
+    purl_only=False,
 ):
     """
     Return Packagedata objects from parsing a Maven pom file at `location` or
@@ -1199,7 +1218,8 @@ def parse(
         package_type=package_type,
         primary_language=primary_language,
         location=location,
-        base_url=base_url
+        base_url=base_url,
+        purl_only=purl_only,
     )
     if package:
         return package
@@ -1212,6 +1232,7 @@ def _parse(
     location=None,
     text=None,
     base_url='https://repo1.maven.org/maven2',
+    purl_only=False,
 ):
     """
     Yield Packagedata objects from parsing a Maven pom file at `location` or
@@ -1249,6 +1270,18 @@ def _parse(
     group_id = pom.group_id
     artifact_id = pom.artifact_id
 
+    pkg = MavenPackageData(
+        datasource_id=datasource_id,
+        type=package_type,
+        namespace=group_id,
+        name=artifact_id,
+        version=version,
+        qualifiers=qualifiers or None,
+    )
+    pkg.dependencies = get_dependencies(pom)
+    if purl_only:
+        return pkg
+
     # craft a source package purl for the main binary
     source_packages = []
     is_main_binary_jar = not classifier and all([group_id, artifact_id, version])
@@ -1283,23 +1316,22 @@ def _parse(
     ))
 
     # FIXME: there are still other data to map in a PackageData
-    return MavenPackageData(
-        datasource_id=datasource_id,
-        type=package_type,
-        primary_language=primary_language,
-        namespace=group_id,
-        name=artifact_id,
-        version=version,
-        qualifiers=qualifiers or None,
-        description=description or None,
-        homepage_url=pom.url or None,
-        extracted_license_statement=extracted_license_statement or None,
-        parties=get_parties(pom),
-        dependencies=get_dependencies(pom),
-        source_packages=source_packages,
-        bug_tracking_url=bug_tracking_url,
-        **urls,
-    )
+    pkg.primary_language = primary_language
+    pkg.description = description or None
+    pkg.homepage_url = pom.url or None
+    pkg.extracted_license_statement = extracted_license_statement or None
+    pkg.parties = get_parties(pom)
+    pkg.source_packages = source_packages
+    pkg.bug_tracking_url = bug_tracking_url
+    (
+        pkg.vcs_url,
+        pkg.code_view_url,
+        pkg.repository_homepage_url,
+        pkg.repository_download_url,
+        pkg.api_data_url,
+    ) = urls
+    return pkg
+
 
 class MavenPackageData(models.PackageData):
 
diff --git a/src/packagedcode/msi.py b/src/packagedcode/msi.py
index 31deea3040b..16d97e24396 100644
--- a/src/packagedcode/msi.py
+++ b/src/packagedcode/msi.py
@@ -124,10 +124,23 @@ def create_package_data_from_msiinfo_results(
     msiinfo_results,
     datasource_id='msi_installer',
     package_type='msi',
+    purl_only=False,
 ):
     """
     Return PackageData from a mapping of `msiinfo_results`
     """
+    subject = msiinfo_results.pop('Subject', '')
+    name = subject
+    version = get_version_from_subject_line(subject)
+    pkg = models.PackageData(
+        datasource_id=datasource_id,
+        type=package_type,
+        name=name,
+        version=version,
+    )
+    if purl_only:
+        return pkg
+
     author_name = msiinfo_results.pop('Author', '')
     parties = []
     if author_name:
@@ -144,27 +157,18 @@ def create_package_data_from_msiinfo_results(
     # the time. Getting the version out of the `Subject` string is not
     # straightforward because the format of the string is usually different
     # between different MSIs
-    subject = msiinfo_results.pop('Subject', '')
-    name = subject
-    version = get_version_from_subject_line(subject)
-    description = msiinfo_results.pop('Comments', '')
-    keywords = msiinfo_results.pop('Keywords', [])
 
-    return models.PackageData(
-        datasource_id=datasource_id,
-        type=package_type,
-        name=name,
-        version=version,
-        description=description,
-        parties=parties,
-        keywords=keywords,
-        extra_data=msiinfo_results
-    )
+    pkg.description = msiinfo_results.pop('Comments', '')
+    pkg.parties = parties
+    pkg.keywords = msiinfo_results.pop('Keywords', [])
+    pkg.extra_data = msiinfo_results
+    return pkg
 
 
 def msi_parse(location,
     datasource_id='msi_installer',
     package_type='msi',
+    purl_only=False,
 ):
     """
     Return PackageData from ``location``
@@ -175,6 +179,7 @@ def msi_parse(location,
             msiinfo_results=info,
             datasource_id=datasource_id,
             package_type=package_type,
+            purl_only=purl_only,
         )
     else:
         return models.PackageData(
@@ -192,5 +197,5 @@ class MsiInstallerHandler(models.DatafileHandler):
     documentation_url = 'https://docs.microsoft.com/en-us/windows/win32/msi/windows-installer-portal'
 
     @classmethod
-    def parse(cls, location):
-        yield msi_parse(location)
+    def parse(cls, location, purl_only=False):
+        yield msi_parse(location=location, purl_only=purl_only)
diff --git a/src/packagedcode/npm.py b/src/packagedcode/npm.py
index 6059850448e..91cdbf61d5f 100644
--- a/src/packagedcode/npm.py
+++ b/src/packagedcode/npm.py
@@ -185,34 +185,55 @@ class NpmPackageJsonHandler(BaseNpmHandler):
     documentation_url = 'https://docs.npmjs.com/cli/v8/configuring-npm/package-json'
 
     @classmethod
-    def _parse(cls, json_data):
+    def _parse(cls, json_data, purl_only=False):
         name = json_data.get('name')
         version = json_data.get('version')
-        homepage_url = json_data.get('homepage', '')
-
-        # a package.json without name and version can be a private package
-
-        if homepage_url and isinstance(homepage_url, list):
-            # TODO: should we keep other URLs
-            homepage_url = homepage_url[0]
-        homepage_url = homepage_url.strip() or None
-
         namespace, name = split_scoped_package_name(name)
 
-        urls = get_urls(namespace, name, version)
         package = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
             namespace=namespace or None,
             name=name,
             version=version or None,
-            description=json_data.get('description', '').strip() or None,
-            homepage_url=homepage_url,
-            **urls,
         )
+        deps_mappers = [
+            ('dependencies', partial(deps_mapper, field_name='dependencies')),
+            ('devDependencies', partial(deps_mapper, field_name='devDependencies')),
+            ('peerDependencies', partial(deps_mapper, field_name='peerDependencies')),
+            ('optionalDependencies', partial(deps_mapper, field_name='optionalDependencies')),
+            ('bundledDependencies', bundle_deps_mapper),
+        ]
+        for source, func in deps_mappers:
+            value = json_data.get(source) or None
+            if value:
+                if isinstance(value, str):
+                    value = value.strip()
+                if value:
+                    func(value, package)
+
+        if purl_only:
+            return package
+
+        homepage_url = json_data.get('homepage', '')
+        # a package.json without name and version can be a private package
+
+        if homepage_url and isinstance(homepage_url, list):
+            # TODO: should we keep other URLs
+            homepage_url = homepage_url[0]
+        homepage_url = homepage_url.strip() or None
+
+        (
+            package.repository_homepage_url,
+            package.repository_download_url,
+            package.api_data_url,
+        ) = get_urls(namespace, name, version)
+        package.primary_language = cls.default_primary_language
+        package.description = json_data.get('description', '').strip() or None
+        package.homepage_url = homepage_url
         vcs_revision = json_data.get('gitHead') or None
 
+
         # mapping of top level package.json items to a function accepting as
         # arguments the package.json element value and returning an iterable of (key,
         # values) to update on a package
@@ -220,12 +241,6 @@ def _parse(cls, json_data):
             ('author', partial(party_mapper, party_type='author')),
             ('contributors', partial(party_mapper, party_type='contributor')),
             ('maintainers', partial(party_mapper, party_type='maintainer')),
-
-            ('dependencies', partial(deps_mapper, field_name='dependencies')),
-            ('devDependencies', partial(deps_mapper, field_name='devDependencies')),
-            ('peerDependencies', partial(deps_mapper, field_name='peerDependencies')),
-            ('optionalDependencies', partial(deps_mapper, field_name='optionalDependencies')),
-            ('bundledDependencies', bundle_deps_mapper),
             ('repository', partial(vcs_repository_mapper, vcs_revision=vcs_revision)),
             ('keywords', keywords_mapper,),
             ('bugs', bugs_mapper),
@@ -257,17 +272,17 @@ def _parse(cls, json_data):
         return package
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with io.open(location, encoding='utf-8') as loc:
             json_data = json.load(loc)
 
-        yield cls._parse(json_data)
+        yield cls._parse(json_data=json_data, purl_only=purl_only)
 
 
 class BaseNpmLockHandler(BaseNpmHandler):
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
 
         with io.open(location, encoding='utf-8') as loc:
             package_data = json.load(loc)
@@ -278,17 +293,13 @@ def parse(cls, location):
         root_version = package_data.get('version')
         root_ns, _ , root_name = root_name.rpartition('/')
 
-        extra_data = dict(lockfile_version=lockfile_version)
         # this is the top level element that we return
         root_package_data = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
             namespace=root_ns,
             name=root_name,
             version=root_version,
-            extra_data=extra_data,
-            **get_urls(root_ns, root_name, root_version)
         )
 
         # https://docs.npmjs.com/cli/v8/configuring-npm/package-lock-json#lockfileversion
@@ -299,7 +310,6 @@ def parse(cls, location):
             deps_key = 'packages'
 
         deps_mapping = package_data.get(deps_key) or {}
-
         dependencies = []
 
         for dep, dep_data in deps_mapping.items():
@@ -359,19 +369,29 @@ def parse(cls, location):
             integrity = dep_data.get('integrity')
             misc.update(get_algo_hexsum(integrity).items())
 
-            resolved_package = models.PackageData(
-                datasource_id=cls.datasource_id,
-                type=cls.default_package_type,
-                primary_language=cls.default_primary_language,
-                namespace=ns,
-                name=name,
-                version=version,
-                extracted_license_statement=extracted_license_statement,
-                **misc,
-            )
-            # these are paths t the root of the installed package in v2
-            if dep:
-                resolved_package.file_references = [models.FileReference(path=dep)],
+            if purl_only:
+                    resolved_package = models.PackageData(
+                    datasource_id=cls.datasource_id,
+                    type=cls.default_package_type,
+                    namespace=ns,
+                    name=name,
+                    version=version,
+                )
+            else:
+                resolved_package = models.PackageData(
+                    datasource_id=cls.datasource_id,
+                    type=cls.default_package_type,
+                    primary_language=cls.default_primary_language,
+                    namespace=ns,
+                    name=name,
+                    version=version,
+                    extracted_license_statement=extracted_license_statement,
+                    **misc,
+                )
+
+                # these are paths t the root of the installed package in v2
+                if dep:
+                    resolved_package.file_references = [models.FileReference(path=dep)],
 
             # v1 as name/constraint pairs
             subrequires = dep_data.get('requires') or {}
@@ -416,6 +436,17 @@ def parse(cls, location):
 
         root_package_data.dependencies = dependencies
 
+        if purl_only:
+            yield root_package_data
+            return
+
+        root_package_data.primary_language = cls.default_primary_language
+        root_package_data.extra_data = dict(lockfile_version=lockfile_version)
+        (
+            root_package_data.repository_homepage_url,
+            root_package_data.repository_download_url,
+            root_package_data.api_data_url,
+        ) = get_urls(root_ns, root_name, root_version)
         yield root_package_data
 
 
@@ -490,7 +521,7 @@ def is_datafile(cls, location, filetypes=tuple()):
         return super().is_datafile(location, filetypes=filetypes) and is_yarn_v2(location)
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         """
         Parse a bew yarn.lock v2 YAML format which looks like this:
 
@@ -545,12 +576,16 @@ def parse(cls, location):
                 )
             top_dependencies.append(dependency)
 
-        yield models.PackageData(
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
             dependencies=top_dependencies,
         )
+        if purl_only:
+            yield pkg
+        else:
+            pkg.primary_language = cls.default_primary_language
+            yield pkg
 
 
 class YarnLockV1Handler(BaseNpmHandler):
@@ -569,7 +604,7 @@ def is_datafile(cls, location, filetypes=tuple()):
         return super().is_datafile(location, filetypes=filetypes) and not is_yarn_v2(location)
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         """
         Parse a classic yarn.lock format which looks like this:
             "@babel/core@^7.1.0", "@babel/core@^7.3.4":
@@ -651,15 +686,24 @@ def parse(cls, location):
             misc.update(get_algo_hexsum(integrity).items())
 
             # we create a resolve package with the details
-            resolved_package_data = models.PackageData(
-                datasource_id=cls.datasource_id,
-                type=cls.default_package_type,
-                namespace=ns,
-                name=name,
-                version=version,
-                primary_language=cls.default_primary_language,
-                **misc,
-            )
+            if purl_only:
+                resolved_package_data = models.PackageData(
+                    datasource_id=cls.datasource_id,
+                    type=cls.default_package_type,
+                    namespace=ns,
+                    name=name,
+                    version=version,
+                )
+            else:
+                resolved_package_data = models.PackageData(
+                    datasource_id=cls.datasource_id,
+                    type=cls.default_package_type,
+                    namespace=ns,
+                    name=name,
+                    version=version,
+                    primary_language=cls.default_primary_language,
+                    **misc,
+                )
 
             # we add the sub-deps to the resolved package
             for subns, subname, subconstraint in sub_dependencies:
@@ -695,12 +739,17 @@ def parse(cls, location):
             )
             dependencies.append(dep)
 
-        yield models.PackageData(
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
             dependencies=dependencies,
         )
+        if purl_only:
+            yield pkg
+        else:
+            pkg.primary_language = cls.default_primary_language
+            yield pkg
+
 
 
 def get_checksum_and_url(url):
diff --git a/src/packagedcode/nuget.py b/src/packagedcode/nuget.py
index dee1afb39d5..adf4a9abe06 100644
--- a/src/packagedcode/nuget.py
+++ b/src/packagedcode/nuget.py
@@ -106,7 +106,7 @@ class NugetNuspecHandler(models.DatafileHandler):
     documentation_url = 'https://docs.microsoft.com/en-us/nuget/reference/nuspec'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with open(location, 'rb') as loc:
             parsed = xmltodict.parse(loc)
 
@@ -121,6 +121,16 @@ def parse(cls, location):
         name = nuspec.get('id')
         version = nuspec.get('version')
 
+        pkg = models.PackageData(
+            datasource_id=cls.datasource_id,
+            type=cls.default_package_type,
+            name=name,
+            version=version,
+        )
+        if purl_only:
+            yield pkg
+            return
+
         # Summary: A short description of the package for UI display. If omitted, a
         # truncated version of description is used.
         description = build_description(nuspec.get('summary'), nuspec.get('description'))
@@ -152,7 +162,11 @@ def parse(cls, location):
             else:
                 vcs_url = vcs_repository
 
-        urls = get_urls(name, version)
+        (
+            pkg.repository_homepage_url,
+            pkg.repository_download_url,
+            pkg.api_data_url,
+        ) = get_urls(name, version)
 
         extracted_license_statement = None
         # See https://docs.microsoft.com/en-us/nuget/reference/nuspec#license
@@ -163,18 +177,11 @@ def parse(cls, location):
         elif 'licenseUrl' in nuspec:
             extracted_license_statement = nuspec.get('licenseUrl')
 
-        yield models.PackageData(
-            datasource_id=cls.datasource_id,
-            type=cls.default_package_type,
-            name=name,
-            version=version,
-            description=description or None,
-            homepage_url=nuspec.get('projectUrl') or None,
-            parties=parties,
-            dependencies=list(get_dependencies(nuspec)),
-            extracted_license_statement=extracted_license_statement,
-            copyright=nuspec.get('copyright') or None,
-            vcs_url=vcs_url,
-            **urls,
-        )
-
+        pkg.description = description or None
+        pkg.homepage_url = nuspec.get('projectUrl') or None
+        pkg.parties = parties
+        pkg.dependencies = list(get_dependencies(nuspec))
+        pkg.extracted_license_statement = extracted_license_statement
+        pkg.copyright = nuspec.get('copyright') or None
+        pkg.vcs_url = vcs_url
+        yield pkg
diff --git a/src/packagedcode/opam.py b/src/packagedcode/opam.py
index 94e2f4942d7..9e6957487e4 100644
--- a/src/packagedcode/opam.py
+++ b/src/packagedcode/opam.py
@@ -31,7 +31,7 @@ def get_package_root(cls, resource, codebase):
         return resource.parent(codebase)
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         opams = parse_opam(location)
 
         package_dependencies = []
@@ -51,24 +51,35 @@ def parse(cls, location):
         name = opams.get('name')
         version = opams.get('version')
 
-        homepage_url = opams.get('homepage')
-        download_url = opams.get('src')
-        vcs_url = opams.get('dev-repo')
-        bug_tracking_url = opams.get('bug-reports')
-        extracted_license_statement = opams.get('license')
-        sha1 = opams.get('sha1')
-        md5 = opams.get('md5')
-        sha256 = opams.get('sha256')
-        sha512 = opams.get('sha512')
-        repository_homepage_url = get_repository_homepage_url(name)
-        api_data_url = get_api_data_url(name, version)
+        pkg = models.PackageData(
+            datasource_id=cls.datasource_id,
+            type=cls.default_package_type,
+            name=name,
+            version=version,
+            dependencies=package_dependencies,
+        )
+        if purl_only:
+            yield pkg
+            return
+
+        pkg.homepage_url = opams.get('homepage')
+        pkg.download_url = opams.get('src')
+        pkg.vcs_url = opams.get('dev-repo')
+        pkg.bug_tracking_url = opams.get('bug-reports')
+        pkg.extracted_license_statement = opams.get('license')
+        pkg.sha1 = opams.get('sha1')
+        pkg.md5 = opams.get('md5')
+        pkg.sha256 = opams.get('sha256')
+        pkg.sha512 = opams.get('sha512')
+        pkg.repository_homepage_url = get_repository_homepage_url(name)
+        pkg.api_data_url = get_api_data_url(name, version)
 
         short_desc = opams.get('synopsis') or ''
         long_desc = opams.get('description') or ''
         if long_desc == short_desc:
             long_desc = None
         descriptions = [d for d in (short_desc, long_desc) if d and d.strip()]
-        description = '\n'.join(descriptions)
+        pkg.description = '\n'.join(descriptions)
 
         parties = []
         authors = opams.get('authors') or []
@@ -90,27 +101,9 @@ def parse(cls, location):
                 )
             )
 
-        yield models.PackageData(
-            datasource_id=cls.datasource_id,
-            type=cls.default_package_type,
-            name=name,
-            version=version,
-            vcs_url=vcs_url,
-            homepage_url=homepage_url,
-            download_url=download_url,
-            sha1=sha1,
-            md5=md5,
-            sha256=sha256,
-            sha512=sha512,
-            bug_tracking_url=bug_tracking_url,
-            extracted_license_statement=extracted_license_statement,
-            description=description,
-            parties=parties,
-            dependencies=package_dependencies,
-            api_data_url=api_data_url,
-            repository_homepage_url=repository_homepage_url,
-            primary_language=cls.default_primary_language
-        )
+        pkg.parties = parties
+        pkg.primary_language = cls.default_primary_language
+        yield pkg
 
     @classmethod
     def assign_package_to_resources(cls, package, resource, codebase, package_adder):
diff --git a/src/packagedcode/phpcomposer.py b/src/packagedcode/phpcomposer.py
index b729e0bfc25..86f7a0b7d45 100644
--- a/src/packagedcode/phpcomposer.py
+++ b/src/packagedcode/phpcomposer.py
@@ -58,7 +58,7 @@ class PhpComposerJsonHandler(BasePhpComposerHandler):
     documentation_url = 'https://getcomposer.org/doc/04-schema.md'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         """
         Yield one or more Package manifest objects given a file ``location``
         pointing to a package archive, manifest or similar.
@@ -69,7 +69,10 @@ def parse(cls, location):
         with io.open(location, encoding='utf-8') as loc:
             package_json = json.load(loc)
 
-        yield build_package_data(package_json)
+        yield build_package_data(
+            package_json=package_json,
+            purl_only=purl_only
+        )
 
 
 def get_repository_homepage_url(namespace, name):
@@ -86,7 +89,7 @@ def get_api_data_url(namespace, name):
         return f'https://packagist.org/p/packages/{name}.json'
 
 
-def build_package_data(package_data):
+def build_package_data(package_data, purl_only=False):
 
     # Note: A composer.json without name and description is not a usable PHP
     # composer package. Name and description fields are required but only for
@@ -108,10 +111,13 @@ def build_package_data(package_data):
         type=PhpComposerJsonHandler.default_package_type,
         namespace=ns,
         name=name,
-        repository_homepage_url=get_repository_homepage_url(ns, name),
-        api_data_url=get_api_data_url(ns, name),
-        primary_language=PhpComposerJsonHandler.default_primary_language,
     )
+    if purl_only:
+        return package
+
+    package.repository_homepage_url = get_repository_homepage_url(ns, name)
+    package.api_data_url = get_api_data_url(ns, name)
+    package.primary_language = PhpComposerJsonHandler.default_primary_language
 
     # mapping of top level composer.json items to the Package object field name
     plain_fields = [
@@ -170,16 +176,16 @@ class PhpComposerLockHandler(BasePhpComposerHandler):
     documentation_url = 'https://getcomposer.org/doc/01-basic-usage.md#commit-your-composer-lock-file-to-version-control'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with io.open(location, encoding='utf-8') as loc:
             package_data = json.load(loc)
 
         packages = [
-            build_package_data(p)
+            build_package_data(package_data=p, purl_only=purl_only)
             for p in package_data.get('packages', [])
         ]
         packages_dev = [
-            build_package_data(p)
+            build_package_data(package_data=p, purl_only=purl_only)
             for p in package_data.get('packages-dev', [])
         ]
 
@@ -192,12 +198,16 @@ def parse(cls, location):
             for p in packages_dev
         ]
 
-        yield models.PackageData(
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
             dependencies=required_deps + required_dev_deps
         )
+        if purl_only:
+            yield pkg
+        else:
+            pkg.primary_language = cls.default_primary_language
+            yield pkg
 
         for package in packages + packages_dev:
             yield package
diff --git a/src/packagedcode/plugin_package.py b/src/packagedcode/plugin_package.py
index 2e1e4ad6e07..c994a3d762e 100644
--- a/src/packagedcode/plugin_package.py
+++ b/src/packagedcode/plugin_package.py
@@ -159,7 +159,16 @@ class PackageScanner(ScanPlugin):
             help_group=SCAN_GROUP,
             sort_order=21,
         ),
-
+        PluggableCommandLineOption(
+            (
+                '--purl',
+            ),
+            is_flag=True,
+            default=False,
+            help='Only detect PURL fields in application package and dependency manifests, lockfiles and related data.',
+            help_group=SCAN_GROUP,
+            sort_order=22,
+        ),
         PluggableCommandLineOption(
             ('--list-packages',),
             is_flag=True,
@@ -170,10 +179,10 @@ class PackageScanner(ScanPlugin):
         ),
     ]
 
-    def is_enabled(self, package, system_package, **kwargs):
-        return package or system_package
+    def is_enabled(self, package, system_package, purl, **kwargs):
+        return package or system_package or purl
 
-    def get_scanner(self, package=True, system_package=False, **kwargs):
+    def get_scanner(self, package=True, system_package=False, purl=False, **kwargs):
         """
         Return a scanner callable to scan a file for package data.
         """
@@ -183,9 +192,10 @@ def get_scanner(self, package=True, system_package=False, **kwargs):
             get_package_data,
             application=package,
             system=system_package,
+            purl_only=purl,
         )
 
-    def process_codebase(self, codebase, strip_root=False, **kwargs):
+    def process_codebase(self, codebase, strip_root=False, purl=False, **kwargs):
         """
         Populate the ``codebase`` top level ``packages`` and ``dependencies``
         with package and dependency instances, assembling parsed package data
@@ -194,6 +204,11 @@ def process_codebase(self, codebase, strip_root=False, **kwargs):
         Also perform additional package license detection that depends on either
         file license detection or the package detections.
         """
+        # If we only want purls, we want to skip both the package
+        # assembly and the extra package license detection steps
+        if purl:
+            return
+
         has_licenses = hasattr(codebase.root, 'license_detections')
 
         # These steps add proper license detections to package_data and hence
diff --git a/src/packagedcode/pubspec.py b/src/packagedcode/pubspec.py
index a6abd5b8642..34285f0202a 100644
--- a/src/packagedcode/pubspec.py
+++ b/src/packagedcode/pubspec.py
@@ -60,11 +60,14 @@ class DartPubspecYamlHandler(BaseDartPubspecHandler):
     documentation_url = 'https://dart.dev/tools/pub/pubspec'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with open(location) as inp:
             pubspec_data = saneyaml.load(inp.read())
 
-        package_data = build_package(pubspec_data)
+        package_data = build_package(
+            pubspec_data=pubspec_data,
+            purl_only=purl_only
+        )
         if package_data:
             yield package_data
 
@@ -78,18 +81,22 @@ class DartPubspecLockHandler(BaseDartPubspecHandler):
     documentation_url = 'https://web.archive.org/web/20220330081004/https://gpalma.pt/blog/what-is-the-pubspec-lock/'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with open(location) as inp:
             locks_data = saneyaml.load(inp.read())
 
         dependencies = list(collect_locks(locks_data))
 
-        yield models.PackageData(
+        pkg = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
             dependencies=dependencies
         )
+        if purl_only:
+            yield pkg
+        else:
+            pkg.primary_language = cls.default_primary_language
+            yield pkg
 
 
 def collect_locks(locks_data):
@@ -238,44 +245,12 @@ def build_dep(name, version, scope, is_runtime=True, is_optional=False):
     return dep
 
 
-def build_package(pubspec_data):
+def build_package(pubspec_data, purl_only=False):
     """
     Return a package object from a package data mapping or None
     """
     name = pubspec_data.get('name')
     version = pubspec_data.get('version')
-    description = pubspec_data.get('description')
-    homepage_url = pubspec_data.get('homepage')
-    extracted_license_statement = pubspec_data.get('license')
-    vcs_url = pubspec_data.get('repository')
-    download_url = pubspec_data.get('archive_url')
-
-    api_data_url = name and version and f'https://pub.dev/api/packages/{name}/versions/{version}'
-    repository_homepage_url = name and version and f'https://pub.dev/packages/{name}/versions/{version}'
-
-    # A URL should be in the form of:
-    # https://pub.dartlang.org/packages/url_launcher/versions/6.0.9.tar.gz
-    # And it may resolve to:
-    # https://storage.googleapis.com/pub-packages/packages/http-0.13.2.tar.gz
-    # as seen in the pub.dev web pages
-    repository_download_url = name and version and f'https://pub.dartlang.org/packages/{name}/versions/{version}.tar.gz'
-
-    download_url = download_url or repository_download_url
-
-    # Author and authors are deprecated
-    authors = []
-    author = pubspec_data.get('author')
-    if author:
-        authors.append(author)
-    authors.extend(pubspec_data.get('authors') or [])
-
-    parties = []
-    for auth  in authors:
-        parties.append(models.Party(
-            type=models.party_person,
-            role='author',
-            name=auth
-        ))
 
     package_dependencies = []
     dependencies = collect_deps(
@@ -302,6 +277,51 @@ def build_package(pubspec_data):
     )
     package_dependencies.extend(env_dependencies)
 
+    package = models.PackageData(
+        datasource_id=DartPubspecYamlHandler.datasource_id,
+        type=DartPubspecYamlHandler.default_primary_language,
+        name=name,
+        version=version,
+        dependencies=package_dependencies,
+    )
+    if purl_only:
+        return package
+
+    package.primary_language = DartPubspecYamlHandler.default_primary_language
+    package.description = pubspec_data.get('description')
+    package.homepage_url = pubspec_data.get('homepage')
+    package.extracted_license_statement = pubspec_data.get('license')
+    package.vcs_url = pubspec_data.get('repository')
+    package.download_url = pubspec_data.get('archive_url')
+
+    package.api_data_url = name and version and f'https://pub.dev/api/packages/{name}/versions/{version}'
+    package.repository_homepage_url = name and version and f'https://pub.dev/packages/{name}/versions/{version}'
+
+    # A URL should be in the form of:
+    # https://pub.dartlang.org/packages/url_launcher/versions/6.0.9.tar.gz
+    # And it may resolve to:
+    # https://storage.googleapis.com/pub-packages/packages/http-0.13.2.tar.gz
+    # as seen in the pub.dev web pages
+    package.repository_download_url = name and version and f'https://pub.dartlang.org/packages/{name}/versions/{version}.tar.gz'
+    package.download_url = package.download_url or package.repository_download_url
+
+    # Author and authors are deprecated
+    authors = []
+    author = pubspec_data.get('author')
+    if author:
+        authors.append(author)
+    authors.extend(pubspec_data.get('authors') or [])
+
+    parties = []
+    for auth  in authors:
+        parties.append(models.Party(
+            type=models.party_person,
+            role='author',
+            name=auth
+        ))
+
+    package.parties = parties
+
     extra_data = {}
 
     def add_to_extra_if_present(_key):
@@ -314,22 +334,6 @@ def add_to_extra_if_present(_key):
     add_to_extra_if_present('dependencies_overrides')
     add_to_extra_if_present('executables')
     add_to_extra_if_present('publish_to')
+    package.extra_data = extra_data
 
-    return models.PackageData(
-        datasource_id=DartPubspecYamlHandler.datasource_id,
-        type=DartPubspecYamlHandler.default_primary_language,
-        primary_language=DartPubspecYamlHandler.default_primary_language,
-        name=name,
-        version=version,
-        download_url=download_url,
-        vcs_url=vcs_url,
-        description=description,
-        extracted_license_statement=extracted_license_statement,
-        parties=parties,
-        homepage_url=homepage_url,
-        dependencies=package_dependencies,
-        extra_data=extra_data,
-        repository_homepage_url=repository_homepage_url,
-        api_data_url=api_data_url,
-        repository_download_url=repository_download_url,
-    )
+    return package
diff --git a/src/packagedcode/pypi.py b/src/packagedcode/pypi.py
index cde769d4de1..b80334e4db3 100644
--- a/src/packagedcode/pypi.py
+++ b/src/packagedcode/pypi.py
@@ -79,11 +79,12 @@ class PythonEggPkgInfoFile(models.DatafileHandler):
     documentation_url = 'https://peps.python.org/pep-0376/'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         yield parse_metadata(
             location=location,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -103,11 +104,12 @@ class PythonEditableInstallationPkgInfoFile(models.DatafileHandler):
     documentation_url = 'https://peps.python.org/pep-0376/'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         yield parse_metadata(
             location=location,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -320,11 +322,12 @@ def is_datafile(cls, location):
         )
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         yield parse_metadata(
             location=location,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
 
@@ -337,11 +340,12 @@ class PythonInstalledWheelMetadataFile(models.DatafileHandler):
     documentation_url = 'https://packaging.python.org/en/latest/specifications/core-metadata/'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         yield parse_metadata(
             location=location,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -457,7 +461,7 @@ class PyprojectTomlHandler(models.NonAssemblableDatafileHandler):
 META_DIR_SUFFIXES = '.dist-info', '.egg-info', 'EGG-INFO',
 
 
-def parse_metadata(location, datasource_id, package_type):
+def parse_metadata(location, datasource_id, package_type, purl_only=False):
     """
     Return a PackageData object from a PKG-INFO or METADATA file at ``location``
     which is a path string or pathlib.Path-like object (including a possible zip
@@ -480,28 +484,34 @@ def parse_metadata(location, datasource_id, package_type):
 
     name = get_attribute(meta, 'Name')
     version = get_attribute(meta, 'Version')
-
-    urls, extra_data = get_urls(metainfo=meta, name=name, version=version)
-
     dependencies = get_dist_dependencies(dist)
 
-    file_references = list(get_file_references(dist))
-
-    return models.PackageData(
-        datasource_id=datasource_id,
-        type=package_type,
-        primary_language='Python',
-        name=name,
-        version=version,
-        extracted_license_statement=get_declared_license(meta),
-        description=get_description(metainfo=meta, location=str(location)),
-        keywords=get_keywords(meta),
-        parties=get_parties(meta),
-        dependencies=dependencies,
-        file_references=file_references,
-        extra_data=extra_data,
-        **urls,
-    )
+    if purl_only:
+        return models.PackageData(
+            datasource_id=datasource_id,
+            type=package_type,
+            name=name,
+            version=version,
+            dependencies=dependencies,
+        )
+    else:
+        urls, extra_data = get_urls(metainfo=meta, name=name, version=version)
+        file_references = list(get_file_references(dist))
+        return models.PackageData(
+            datasource_id=datasource_id,
+            type=package_type,
+            primary_language='Python',
+            name=name,
+            version=version,
+            dependencies=dependencies,
+            extracted_license_statement=get_declared_license(meta),
+            description=get_description(metainfo=meta, location=str(location)),
+            keywords=get_keywords(meta),
+            parties=get_parties(meta),
+            file_references=file_references,
+            extra_data=extra_data,
+            **urls,
+        )
 
 
 def urlsafe_b64decode(data):
@@ -551,7 +561,7 @@ class PypiWheelHandler(models.DatafileHandler):
     documentation_url = 'https://peps.python.org/pep-0427/'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with zipfile.ZipFile(location) as zf:
             for path in ZipPath(zf).iterdir():
                 if not path.name.endswith(META_DIR_SUFFIXES):
@@ -564,6 +574,7 @@ def parse(cls, location):
                         location=metapath,
                         datasource_id=cls.datasource_id,
                         package_type=cls.default_package_type,
+                        purl_only=purl_only,
                     )
 
 
@@ -577,7 +588,7 @@ class PypiEggHandler(models.DatafileHandler):
     documentation_url = 'https://web.archive.org/web/20210604075235/http://peak.telecommunity.com/DevCenter/PythonEggs'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with zipfile.ZipFile(location) as zf:
             for path in ZipPath(zf).iterdir():
                 if not path.name.endswith(META_DIR_SUFFIXES):
@@ -591,6 +602,7 @@ def parse(cls, location):
                         location=metapath,
                         datasource_id=cls.datasource_id,
                         package_type=cls.default_package_type,
+                        purl_only=purl_only,
                     )
 
 
@@ -610,7 +622,7 @@ def is_datafile(cls, location, filetypes=tuple()):
             return True
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         # FIXME: add dependencies
 
         try:
@@ -622,19 +634,27 @@ def parse(cls, location):
         version = sdist.version
         urls, extra_data = get_urls(metainfo=sdist, name=name, version=version)
 
-        yield models.PackageData(
-            datasource_id=cls.datasource_id,
-            type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
-            name=name,
-            version=version,
-            description=get_description(sdist, location=location),
-            extracted_license_statement=get_declared_license(sdist),
-            keywords=get_keywords(sdist),
-            parties=get_parties(sdist),
-            extra_data=extra_data,
-            **urls,
-        )
+        if purl_only:
+            yield models.PackageData(
+                datasource_id=cls.datasource_id,
+                type=cls.default_package_type,
+                name=name,
+                version=version,
+            )
+        else:
+            yield models.PackageData(
+                datasource_id=cls.datasource_id,
+                type=cls.default_package_type,
+                primary_language=cls.default_primary_language,
+                name=name,
+                version=version,
+                description=get_description(sdist, location=location),
+                extracted_license_statement=get_declared_license(sdist),
+                keywords=get_keywords(sdist),
+                parties=get_parties(sdist),
+                extra_data=extra_data,
+                **urls,
+            )
 
 
 class PythonSetupPyHandler(BaseExtractedPythonLayout):
@@ -646,38 +666,46 @@ class PythonSetupPyHandler(BaseExtractedPythonLayout):
     documentation_url = 'https://docs.python.org/3/distutils/setupscript.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         setup_args = get_setup_py_args(location)
 
         # it may be legit to have a name-less package?
         # in anycase we do not want to fail because of that
         name = setup_args.get('name')
-
         version = setup_args.get('version')
         if not version:
             # search for possible dunder versions here and elsewhere
             version = detect_version_attribute(location)
 
-        urls, extra_data = get_urls(metainfo=setup_args, name=name, version=version)
-
         dependencies = get_setup_py_dependencies(setup_args)
-        python_requires = get_setup_py_python_requires(setup_args)
-        extra_data.update(python_requires)
 
-        yield models.PackageData(
-            datasource_id=cls.datasource_id,
-            type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
-            name=name,
-            version=version,
-            description=get_description(setup_args),
-            parties=get_setup_parties(setup_args),
-            extracted_license_statement=get_declared_license(setup_args),
-            dependencies=dependencies,
-            keywords=get_keywords(setup_args),
-            extra_data=extra_data,
-            **urls,
-        )
+        if purl_only:
+            yield models.PackageData(
+                datasource_id=cls.datasource_id,
+                type=cls.default_package_type,
+                name=name,
+                version=version,
+                dependencies=dependencies,
+            )
+        else:
+            urls, extra_data = get_urls(metainfo=setup_args, name=name, version=version)
+            python_requires = get_setup_py_python_requires(setup_args)
+            extra_data.update(python_requires)
+
+            yield models.PackageData(
+                datasource_id=cls.datasource_id,
+                type=cls.default_package_type,
+                primary_language=cls.default_primary_language,
+                name=name,
+                version=version,
+                description=get_description(setup_args),
+                parties=get_setup_parties(setup_args),
+                extracted_license_statement=get_declared_license(setup_args),
+                dependencies=dependencies,
+                keywords=get_keywords(setup_args),
+                extra_data=extra_data,
+                **urls,
+            )
 
 
 class ResolvedPurl(NamedTuple):
@@ -694,7 +722,7 @@ class BaseDependencyFileHandler(models.DatafileHandler):
     """
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         file_name = fileutils.file_name(location)
 
         dependency_type = get_dparse2_supported_file_name(file_name)
@@ -705,12 +733,19 @@ def parse(cls, location):
             location=location,
             file_name=dependency_type,
         )
-        yield models.PackageData(
-            datasource_id=cls.datasource_id,
-            type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
-            dependencies=dependencies,
-        )
+        if purl_only:
+            yield models.PackageData(
+                datasource_id=cls.datasource_id,
+                type=cls.default_package_type,
+                dependencies=dependencies,
+            )
+        else:
+            yield models.PackageData(
+                datasource_id=cls.datasource_id,
+                type=cls.default_package_type,
+                primary_language=cls.default_primary_language,
+                dependencies=dependencies,
+            )
 
 
 class SetupCfgHandler(BaseExtractedPythonLayout):
@@ -722,7 +757,7 @@ class SetupCfgHandler(BaseExtractedPythonLayout):
     documentation_url = 'https://peps.python.org/pep-0390/'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
 
         metadata = {}
         parser = ConfigParser()
@@ -781,35 +816,45 @@ def parse(cls, location):
                     if not content:
                         continue
                     metadata[name] = content
-
-        parties = []
-        author = metadata.get('author')
-        if author:
-            parties = [
-                models.Party(
-                    type=models.party_person,
-                    name=author,
-                    role='author',
-                    email=metadata.get('author_email'),
-                )
-            ]
-
-        extracted_license_statement = metadata.get('license')
-        license_file_references = metadata.get('license_files')
-        if license_file_references:
-            extracted_license_statement = f"{extracted_license_statement} {license_file_references}"
-
-        yield models.PackageData(
+        
+        package = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
             name=metadata.get('name'),
             version=metadata.get('version'),
-            parties=parties,
-            homepage_url=metadata.get('url'),
-            primary_language=cls.default_primary_language,
             dependencies=dependent_packages,
-            extracted_license_statement=extracted_license_statement,
         )
+        if purl_only:
+            yield package
+        else:
+            parties = []
+            author = metadata.get('author')
+            if author:
+                parties = [
+                    models.Party(
+                        type=models.party_person,
+                        name=author,
+                        role='author',
+                        email=metadata.get('author_email'),
+                    )
+                ]
+
+            extracted_license_statement = metadata.get('license')
+            license_file_references = metadata.get('license_files')
+            if license_file_references:
+                extracted_license_statement = f"{extracted_license_statement} {license_file_references}"
+
+            yield models.PackageData(
+                datasource_id=cls.datasource_id,
+                type=cls.default_package_type,
+                name=metadata.get('name'),
+                version=metadata.get('version'),
+                parties=parties,
+                homepage_url=metadata.get('url'),
+                primary_language=cls.default_primary_language,
+                dependencies=dependent_packages,
+                extracted_license_statement=extracted_license_statement,
+            )
 
     @classmethod
     def parse_reqs(cls, reqs, scope):
@@ -871,7 +916,7 @@ class PipfileLockHandler(BaseDependencyFileHandler):
     documentation_url = 'https://github.com/pypa/pipfile'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with open(location) as f:
             content = f.read()
 
@@ -888,13 +933,17 @@ def parse(cls, location):
             file_name='Pipfile.lock',
         )
 
-        yield models.PackageData(
+        package = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
-            sha256=sha256,
             dependencies=dependent_packages,
         )
+        if purl_only:
+            yield package
+        else:
+            package.primary_language = cls.default_primary_language
+            package.sha256 = sha256
+            yield package
 
 
 class PipRequirementsFileHandler(BaseDependencyFileHandler):
@@ -917,15 +966,19 @@ class PipRequirementsFileHandler(BaseDependencyFileHandler):
     documentation_url = 'https://pip.pypa.io/en/latest/reference/requirements-file-format/'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         dependencies, extra_data = get_requirements_txt_dependencies(location=location)
-        yield models.PackageData(
+        package = models.PackageData(
             datasource_id=cls.datasource_id,
             type=cls.default_package_type,
-            primary_language=cls.default_primary_language,
             dependencies=dependencies,
-            extra_data=extra_data,
         )
+        if purl_only:
+            yield package
+        else:
+            package.primary_language = cls.default_primary_language
+            package.extra_data = extra_data
+            yield package
 
 # TODO: enable nested load
 
diff --git a/src/packagedcode/readme.py b/src/packagedcode/readme.py
index 63741e86648..4cf4a9a8c69 100644
--- a/src/packagedcode/readme.py
+++ b/src/packagedcode/readme.py
@@ -55,11 +55,14 @@ class ReadmeHandler(models.NonAssemblableDatafileHandler):
     documentation_url = ''
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with open(location, encoding='utf-8') as loc:
             readme_manifest = loc.read()
 
-        package_data = build_package(readme_manifest)
+        package_data = build_package(
+            readme_manifest=readme_manifest,
+            purl_only=purl_only
+        )
 
         if not package_data.name:
             # If no name was detected for the Package, then we use the basename
@@ -71,7 +74,7 @@ def parse(cls, location):
         yield package_data
 
 
-def build_package(readme_manifest):
+def build_package(readme_manifest, purl_only=False):
     """
     Return a Package object from a readme_manifest mapping (from a
     README.chromium file or similar) or None.
@@ -102,7 +105,11 @@ def build_package(readme_manifest):
         package_key = PACKAGE_FIELD_BY_README_FIELD.get(key)
         if not package_key:
             continue
+        if purl_only and package_key not in ["name", "version"]:
+            continue
+
         setattr(package, package_key, value)
 
-    package.populate_license_fields()
+    if not purl_only:
+        package.populate_license_fields()
     return package
diff --git a/src/packagedcode/recognize.py b/src/packagedcode/recognize.py
index c7e794ecaf6..e0dc315cb30 100644
--- a/src/packagedcode/recognize.py
+++ b/src/packagedcode/recognize.py
@@ -44,6 +44,7 @@ def recognize_package_data(
     location,
     application=True,
     system=False,
+    purl_only=False,
 ):
     """
     Return a list of Package objects if any package_data were recognized for
@@ -63,12 +64,19 @@ def recognize_package_data(
     elif system:
         datafile_handlers = SYSTEM_PACKAGE_DATAFILE_HANDLERS
 
-    return list(_parse(location, datafile_handlers=datafile_handlers))
+    return list(
+        _parse(
+            location,
+            datafile_handlers=datafile_handlers,
+            purl_only=purl_only,
+        )
+    )
 
 
 def _parse(
     location,
     datafile_handlers=APPLICATION_PACKAGE_DATAFILE_HANDLERS,
+    purl_only=False,
 ):
     """
     Yield parsed PackageData objects from ``location``. Raises Exceptions on errors.
@@ -85,7 +93,7 @@ def _parse(
             logger_debug(f'_parse:.is_datafile: {location}')
 
         try:
-            for parsed in handler.parse(location):
+            for parsed in handler.parse(location, purl_only=purl_only):
                 if TRACE:
                     logger_debug(f' _parse: parsed: {parsed!r}')
                 yield parsed
diff --git a/src/packagedcode/rpm.py b/src/packagedcode/rpm.py
index 01ae1ab850e..cc4e8bccb27 100644
--- a/src/packagedcode/rpm.py
+++ b/src/packagedcode/rpm.py
@@ -121,7 +121,7 @@ def to_string(self):
 class BaseRpmInstalledDatabaseHandler(models.DatafileHandler):
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         # we receive the location of the Package database file and we need to
         # scan the parent which is the directory that contains the rpmdb
         loc_path = Path(location)
@@ -133,6 +133,7 @@ def parse(cls, location):
             location=xmlish_loc,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
         # TODO: package_data.namespace = cls.default_package_namespace
         return package_data
@@ -272,7 +273,7 @@ class RpmArchiveHandler(models.DatafileHandler):
     documentation_url = 'https://en.wikipedia.org/wiki/RPM_Package_Manager'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         rpm_tags = get_rpm_tags(location, include_desc=True)
 
         if TRACE: logger_debug('recognize: rpm_tags', rpm_tags)
@@ -319,6 +320,18 @@ def parse(cls, location):
             if TRACE: logger_debug('recognize: source_rpm', src_purl)
             source_packages = [src_purl]
 
+        package = models.PackageData(
+            datasource_id=cls.datasource_id,
+            type=cls.default_package_type,
+            # TODO: namespace=cls.default_package_namespace,
+            name=name,
+            version=evr,
+            source_packages=source_packages,
+        )
+        if purl_only:
+            yield package
+            return
+
         parties = []
 
         # TODO: also use me to craft a namespace!!!
@@ -335,13 +348,13 @@ def parse(cls, location):
         if rpm_tags.vendor:
             parties.append(models.Party(name=rpm_tags.vendor, role='vendor'))
 
-        description = build_description(summary=rpm_tags.summary, description=rpm_tags.description)
+        package.description = build_description(summary=rpm_tags.summary, description=rpm_tags.description)
 
         if TRACE:
             data = dict(
                 name=name,
                 version=evr,
-                description=description or None,
+                description=package.description or None,
                 homepage_url=rpm_tags.url or None,
                 parties=parties,
                 extracted_license_statement=rpm_tags.license or None,
@@ -349,18 +362,9 @@ def parse(cls, location):
             )
             logger_debug('recognize: data to create a package:\n', data)
 
-        package = models.PackageData(
-            datasource_id=cls.datasource_id,
-            type=cls.default_package_type,
-            # TODO: namespace=cls.default_package_namespace,
-            name=name,
-            version=evr,
-            description=description or None,
-            homepage_url=rpm_tags.url or None,
-            parties=parties,
-            extracted_license_statement=rpm_tags.license or None,
-            source_packages=source_packages,
-        )
+        package.homepage_url = rpm_tags.url or None
+        package.parties = parties
+        package.extracted_license_statement = rpm_tags.license or None
 
         if TRACE:
             logger_debug('recognize: created package:\n', package)
diff --git a/src/packagedcode/rpm_installed.py b/src/packagedcode/rpm_installed.py
index 5d01436d2e1..aa66d32c924 100644
--- a/src/packagedcode/rpm_installed.py
+++ b/src/packagedcode/rpm_installed.py
@@ -36,7 +36,7 @@ def logger_debug(*args):
         return logger.debug(' '.join(isinstance(a, str) and a or repr(a) for a in args))
 
 
-def parse_rpm_xmlish(location, datasource_id, package_type):
+def parse_rpm_xmlish(location, datasource_id, package_type, purl_only=False):
     """
     Yield PackageData built from an RPM XML'ish file at ``location``. This is a file
     created with the rpm CLI with the xml query option.
@@ -58,6 +58,7 @@ def parse_rpm_xmlish(location, datasource_id, package_type):
             rpm_tags=tags,
             datasource_id=datasource_id,
             package_type=package_type,
+            purl_only=purl_only,
         )
 
 
@@ -133,7 +134,13 @@ def collect_tags(raw_tags):
         yield name, value_type, value
 
 
-def build_package(rpm_tags, datasource_id, package_type, package_namespace=None):
+def build_package(
+    rpm_tags,
+    datasource_id,
+    package_type,
+    package_namespace=None,
+    purl_only=False,
+):
     """
     Return a PackageData object from an ``rpm_tags`` iterable of (name,
     value_type, value) tuples.
@@ -147,7 +154,10 @@ def build_package(rpm_tags, datasource_id, package_type, package_namespace=None)
     }
 
     for name, _value_type, value in rpm_tags:
-        handler = RPM_TAG_HANDLER_BY_NAME.get(name)
+        handler = RPM_TAG_HANDLER_BY_NAME_PURLS.get(name)
+        if not handler and not purl_only:
+            handler = RPM_TAG_HANDLER_BY_NAME_OTHERS.get(name)
+
         # FIXME: we need to handle EVRA correctly
         # TODO: add more fields
         # TODO: merge with tag handling in rpm.py
@@ -283,7 +293,7 @@ def dirname_handler(value, **kwargs):
 #   PackageData field name
 
 
-RPM_TAG_HANDLER_BY_NAME = {
+RPM_TAG_HANDLER_BY_NAME_PURLS = {
 
     ############################################################################
     # per-package fields
@@ -294,11 +304,15 @@ def dirname_handler(value, **kwargs):
     #  'Epoch'
     #  'Release' 11.3.2
     'Version': name_value_str_handler('version'),
+    'Arch': arch_handler,
+}
+
+
+RPM_TAG_HANDLER_BY_NAME_OTHERS = {
     'Description': name_value_str_handler('description'),
     'Sha1header': name_value_str_handler('sha1'),
     'Url': name_value_str_handler('homepage_url'),
     'License': name_value_str_handler('extracted_license_statement'),
-    'Arch': arch_handler,
     'Size': size_handler,
 
     # TODO:
diff --git a/src/packagedcode/rubygems.py b/src/packagedcode/rubygems.py
index 6b77d6945d6..1a5834484a5 100644
--- a/src/packagedcode/rubygems.py
+++ b/src/packagedcode/rubygems.py
@@ -41,12 +41,13 @@ class GemArchiveHandler(models.DatafileHandler):
     )
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         metadata = extract_gem_metadata(location)
         metadata = saneyaml.load(metadata)
         yield build_rubygem_package_data(
             gem_data=metadata,
             datasource_id=cls.datasource_id,
+            purl_only=purl_only,
         )
 
 
@@ -84,13 +85,14 @@ class GemMetadataArchiveExtractedHandler(models.DatafileHandler):
     )
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with open(location, 'rb') as met:
             metadata = met.read()
         metadata = saneyaml.load(metadata)
         yield build_rubygem_package_data(
             gem_data=metadata,
             datasource_id=cls.datasource_id,
+            purl_only=purl_only,
         )
 
     @classmethod
@@ -129,7 +131,7 @@ class GemspecHandler(models.DatafileHandler):
     documentation_url = 'https://guides.rubygems.org/specification-reference/'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         gemspec = spec.parse_spec(
             location=location,
             package_type=cls.default_package_type,
@@ -137,6 +139,19 @@ def parse(cls, location):
 
         name = gemspec.get('name')
         version = gemspec.get('version')
+        dependencies = gemspec.get('dependencies') or []
+
+        package = models.PackageData(
+            datasource_id=cls.datasource_id,
+            type=cls.default_package_type,
+            name=name,
+            version=version,
+            dependencies=dependencies,
+        )
+        if purl_only:
+            yield package
+            return
+
         homepage_url = gemspec.get('homepage')
 
         description = build_description(
@@ -148,24 +163,22 @@ def parse(cls, location):
         extracted_license_statement = gemspec.get('license')
 
         parties = get_parties(gemspec)
-        dependencies = gemspec.get('dependencies') or []
 
-        urls = get_urls(name=name, version=version)
+        (
+            package.repository_homepage_url,
+            package.repository_download_url,
+            package.api_data_url,
+            package.download_url,
+        ) = get_urls(name=name, version=version)
 
-        yield models.PackageData(
-            datasource_id=cls.datasource_id,
-            type=cls.default_package_type,
-            name=name,
-            version=version,
-            parties=parties,
-            homepage_url=homepage_url,
-            vcs_url=vcs_url,
-            description=description,
-            extracted_license_statement=extracted_license_statement,
-            primary_language=cls.default_primary_language,
-            dependencies=dependencies,
-            **urls
-        )
+        package.parties = parties
+        package.homepage_url = homepage_url
+        package.vcs_url = vcs_url
+        package.description = description
+        package.extracted_license_statement = extracted_license_statement
+        package.primary_language = cls.default_primary_language
+
+        yield package
 
 class GemspecInExtractedGemHandler(GemspecHandler):
     datasource_id = 'gemspec_extracted'
@@ -234,7 +247,7 @@ class GemfileLockHandler(BaseGemProjectHandler):
     documentation_url = 'https://bundler.io/man/gemfile.5.html'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         gemfile_lock = GemfileLockParser(location)
         all_gems = list(gemfile_lock.all_gems.values())
         if not all_gems:
@@ -257,16 +270,25 @@ def parse(cls, location):
                 ) for dep in all_gems if dep != primary_gem
             ]
             urls = get_urls(primary_gem.name, primary_gem.version)
-
-            yield models.PackageData(
+            
+            package = models.PackageData(
                 datasource_id=cls.datasource_id,
-                primary_language=cls.default_primary_language,
                 type=cls.default_package_type,
                 name=primary_gem.name,
                 version=primary_gem.version,
                 dependencies=deps,
-                **urls
             )
+            if purl_only:
+                yield package
+            else:
+                (
+                    package.repository_homepage_url,
+                    package.repository_download_url,
+                    package.api_data_url,
+                    package.download_url,
+                ) = urls
+                package.primary_language = cls.default_primary_language
+                yield package
         else:
             deps = [
                 models.DependentPackage(
@@ -417,7 +439,7 @@ def extract_gem_metadata(location):
             fileutils.delete(extract_loc)
 
 
-def build_rubygem_package_data(gem_data, datasource_id):
+def build_rubygem_package_data(gem_data, datasource_id, purl_only=False):
     """
     Return a PackageData for ``datasource_id`` built from a Gem `gem_data`
     mapping or None. The ``gem_data`` can come from a .gemspec or .gem/metadata.
@@ -439,6 +461,17 @@ def build_rubygem_package_data(gem_data, datasource_id):
     else:
         qualifiers = {}
 
+    dependencies = get_dependencies(gem_data.get('dependencies'))
+
+    package_data = models.PackageData(
+        datasource_id=datasource_id,
+        type=GemArchiveHandler.default_package_type,
+        name=name,
+        version=version,
+        qualifiers=qualifiers,
+        dependencies=dependencies,
+    )
+
     description = build_description(
         summary=gem_data.get('summary'),
         description=gem_data.get('description'),
@@ -457,26 +490,21 @@ def build_rubygem_package_data(gem_data, datasource_id):
     if not homepage_url:
         homepage_url = gem_data.get('homepage')
 
-    urls = get_urls(name, version, platform)
-    dependencies = get_dependencies(gem_data.get('dependencies'))
+    (
+        package_data.repository_homepage_url,
+        package_data.repository_download_url,
+        package_data.api_data_url,
+        package_data.download_url,
+    ) = get_urls(name, version, platform)
     file_references = get_file_references(metadata.get('files'))
 
-    package_data = models.PackageData(
-        datasource_id=datasource_id,
-        type=GemArchiveHandler.default_package_type,
-        primary_language=GemArchiveHandler.default_primary_language,
-        name=name,
-        version=version,
-        qualifiers=qualifiers,
-        description=description,
-        homepage_url=homepage_url,
-        extracted_license_statement=extracted_license_statement,
-        bug_tracking_url=metadata.get('bug_tracking_uri'),
-        code_view_url=metadata.get('source_code_uri'),
-        file_references=file_references,
-        dependencies=dependencies,
-        **urls,
-    )
+    package_data.primary_language = GemArchiveHandler.default_primary_language
+    package_data.description = description
+    package_data.homepage_url = homepage_url
+    package_data.extracted_license_statement = extracted_license_statement
+    package_data.bug_tracking_url = metadata.get('bug_tracking_uri')
+    package_data.code_view_url = metadata.get('source_code_uri')
+    package_data.file_references = file_references
 
     # we can have one singular or a plural list of authors
     authors = gem_data.get('authors') or []
diff --git a/src/packagedcode/spec.py b/src/packagedcode/spec.py
index c0bf7083c51..49762a3776d 100644
--- a/src/packagedcode/spec.py
+++ b/src/packagedcode/spec.py
@@ -120,9 +120,12 @@ def get_authors(line):
 
 
 # mapping of parser callable by its field name
-PARSER_BY_NAME = {
+PARSER_BY_NAME_PURL_ONLY = {
     'name': partial(get_value, name='name', matcher=parse_name),
     'version': partial(get_value, name='version', matcher=parse_version),
+}
+
+PARSER_BY_NAME_OTHERS = {
     'license': partial(get_value, name='license', matcher=parse_license),
     'summary': partial(get_value, name='summary', matcher=parse_summary),
     'description': partial(get_value, name='description', matcher=parse_description, clean=False),
@@ -133,7 +136,7 @@ def get_authors(line):
 }
 
 
-def parse_spec(location, package_type):
+def parse_spec(location, package_type, purl_only=False):
     """
     Return a mapping of data parsed from a podspec/gemspec/Pofile/Gemfile file
     at ``location``. Use ``package_type`` a Package URL type for dependencies.
@@ -147,10 +150,27 @@ def parse_spec(location, package_type):
     for line in lines:
         line = pre_process(line)
 
-        for attribute_name, parser in PARSER_BY_NAME.items():
+        for attribute_name, parser in PARSER_BY_NAME_PURL_ONLY.items():
             parsed = parser(line=line)
             if parsed:
                 spec_data[attribute_name] = parsed
+        
+        if not purl_only:
+            for attribute_name, parser in PARSER_BY_NAME_PURL_ONLY.items():
+                parsed = parser(line=line)
+                if parsed:
+                    spec_data[attribute_name] = parsed
+
+    # We avoid reloading twice the file but we are still parsing twice: need to
+    # merge all in gemfileparser or write a better parser.
+    spec_data['dependencies'] = list(get_dependent_packages(
+        lines=lines,
+        location=location,
+        package_type=package_type,
+    ))
+
+    if purl_only:
+        return spec_data
 
     # description can be in single or multi-lines
     # There are many different ways to write description.
@@ -167,14 +187,6 @@ def parse_spec(location, package_type):
             # a single quoted description
             spec_data['description'] = get_cleaned_string(description)
 
-    # We avoid reloading twice the file but we are still parsing twice: need to
-    # merge all in gemfileparser or write a better parser.
-    spec_data['dependencies'] = list(get_dependent_packages(
-        lines=lines,
-        location=location,
-        package_type=package_type,
-    ))
-
     return spec_data
 
 
diff --git a/src/packagedcode/win_pe.py b/src/packagedcode/win_pe.py
index a51ceb9496e..ce077ce12ca 100644
--- a/src/packagedcode/win_pe.py
+++ b/src/packagedcode/win_pe.py
@@ -276,7 +276,7 @@ def is_datafile(cls, location, filetypes=tuple()):
             return True
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         infos = pe_info(location)
 
         version = get_first(
@@ -298,6 +298,16 @@ def parse(cls, location):
             'OriginalFilename',
             'InternalName',
         )
+        package = models.PackageData(
+            datasource_id=cls.datasource_id,
+            type=cls.default_package_type,
+            name=name,
+            version=version,
+        )
+        if purl_only:
+            yield package
+            return
+
         copyr = get_first(infos, 'LegalCopyright')
 
         LegalCopyright = copyr,
@@ -328,15 +338,10 @@ def parse(cls, location):
             parties = [Party(type=party_org, role='author', name=cname)]
         homepage_url = get_first(infos, 'URL', 'WWW')
 
-        yield models.PackageData(
-            datasource_id=cls.datasource_id,
-            type=cls.default_package_type,
-            name=name,
-            version=version,
-            release_date=release_date,
-            copyright=copyr,
-            extracted_license_statement=extracted_license_statement,
-            description=description,
-            parties=parties,
-            homepage_url=homepage_url,
-        )
+        package.release_date = release_date
+        package.copyright = copyr
+        package.extracted_license_statement = extracted_license_statement
+        package.description = description
+        package.parties = parties
+        package.homepage_url = homepage_url
+        yield package
diff --git a/src/packagedcode/win_reg.py b/src/packagedcode/win_reg.py
index e4e59cd3b1e..7a6902ce961 100644
--- a/src/packagedcode/win_reg.py
+++ b/src/packagedcode/win_reg.py
@@ -77,6 +77,7 @@ def get_installed_dotnet_versions_from_hive(
     datasource_id,
     package_type,
     registry_path='\\Microsoft\\NET Framework Setup\\NDP',
+    purl_only=False,
 ):
     """
     Yield PackageData for the installed versions of .NET framework from the
@@ -90,6 +91,7 @@ def get_installed_dotnet_versions_from_hive(
         registry_tree=registry_tree,
         datasource_id=datasource_id,
         package_type=package_type,
+        purl_only=purl_only,
     )
 
 
@@ -97,6 +99,7 @@ def get_installed_dotnet_versions_from_regtree(
     registry_tree,
     datasource_id,
     package_type,
+    purl_only=False,
 ):
     """
     Yield PackageData for the installed versions of .NET framework from a
@@ -111,6 +114,16 @@ def get_installed_dotnet_versions_from_regtree(
         if not entry.get('path', '').endswith('\\Full'):
             continue
 
+        package = models.PackageData(
+            datasource_id=datasource_id,
+            type=package_type,
+            name='microsoft-dot-net-framework',
+            version=version,
+        )
+        if purl_only:
+            yield package
+            return
+
         file_references = []
         version = None
         for values in entry.get('values', []):
@@ -122,13 +135,8 @@ def get_installed_dotnet_versions_from_regtree(
             if key == 'InstallPath':
                 file_references.append(models.FileReference(path=value))
 
-        yield models.PackageData(
-            datasource_id=datasource_id,
-            type=package_type,
-            name='microsoft-dot-net-framework',
-            version=version,
-            file_references=file_references,
-        )
+        package.file_references = file_references
+        yield package
 
 
 def get_installed_windows_programs_from_hive(
@@ -136,6 +144,7 @@ def get_installed_windows_programs_from_hive(
     datasource_id,
     package_type,
     registry_path='\\Microsoft\\Windows\\CurrentVersion\\Uninstall',
+    purl_only=False,
 ):
     """
     Yield installed Windows PackageData from a Windows registry file at
@@ -151,6 +160,7 @@ def get_installed_windows_programs_from_hive(
         registry_tree=registry_tree,
         datasource_id=datasource_id,
         package_type=package_type,
+        purl_only=purl_only,
     )
 
 
@@ -158,6 +168,7 @@ def get_installed_windows_programs_from_regtree(
     registry_tree,
     datasource_id,
     package_type,
+    purl_only=False,
 ):
     """
     Yield installed Windows PackageData from a Windows ``registry_tree``.
@@ -187,6 +198,16 @@ def get_installed_windows_programs_from_regtree(
         name = package_info.get('name')
         version = package_info.get('version')
 
+        package = models.PackageData(
+            datasource_id=datasource_id,
+            type=package_type,
+            name=name,
+            version=version,
+        )
+        if purl_only:
+            yield package
+            return
+
         homepage_url = package_info.get('homepage_url')
         publisher = package_info.get('publisher')
 
@@ -213,21 +234,17 @@ def get_installed_windows_programs_from_regtree(
         if uninstall_string:
             file_references.append(models.FileReference(path=uninstall_string))
 
-        yield models.PackageData(
-            datasource_id=datasource_id,
-            type=package_type,
-            name=name,
-            version=version,
-            parties=parties,
-            homepage_url=homepage_url,
-            file_references=file_references,
-        )
+        package.parties = parties
+        package.homepage_url = homepage_url
+        package.file_references = file_references
+        yield package
 
 
 def get_packages_from_registry_from_hive(
     location,
     datasource_id,
     package_type,
+    purl_only=False,
 ):
     """
     Yield PackageData for Installed Windows Programs from the Windows registry
@@ -238,6 +255,7 @@ def get_packages_from_registry_from_hive(
         datasource_id=datasource_id,
         package_type=package_type,
         registry_path='\\Microsoft\\Windows\\CurrentVersion\\Uninstall',
+        purl_only=purl_only,
     )
 
     yield from get_installed_windows_programs_from_hive(
@@ -245,6 +263,7 @@ def get_packages_from_registry_from_hive(
         datasource_id=datasource_id,
         package_type=package_type,
         registry_path='\\Wow6432Node\\Microsoft\\Windows\\CurrentVersion\\Uninstall',
+        purl_only=purl_only,
     )
 
     yield from get_installed_dotnet_versions_from_hive(
@@ -252,6 +271,7 @@ def get_packages_from_registry_from_hive(
         datasource_id=datasource_id,
         package_type=package_type,
         registry_path='\\Microsoft\\NET Framework Setup\\NDP',
+        purl_only=purl_only,
     )
 
 
@@ -342,11 +362,12 @@ class BaseRegInstalledProgramHandler(models.DatafileHandler):
     root_path_relative_to_datafile_path = None
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         yield from get_packages_from_registry_from_hive(
             location=location,
             datasource_id=cls.datasource_id,
             package_type=cls.default_package_type,
+            purl_only=purl_only,
         )
 
     @classmethod
diff --git a/src/packagedcode/windows.py b/src/packagedcode/windows.py
index 09d62f4c2d9..bc03f04b62d 100644
--- a/src/packagedcode/windows.py
+++ b/src/packagedcode/windows.py
@@ -20,22 +20,30 @@ class MicrosoftUpdateManifestHandler(models.NonAssemblableDatafileHandler):
     description = 'Microsoft Update Manifest .mum file'
 
     @classmethod
-    def parse(cls, location):
+    def parse(cls, location, purl_only=False):
         with open(location , 'rb') as loc:
             parsed = xmltodict.parse(loc)
 
         if not parsed:
             return
 
+        package = models.PackageData(
+            datasource_id=cls.datasource_id,
+            type=cls.default_package_type,
+            name=assembly_identity.get('@name', ''),
+            version=assembly_identity.get('@version', ''),
+        )
+        if purl_only:
+            yield package
+            return
+
         assembly = parsed.get('assembly', {})
-        description = assembly.get('@description', '')
+        package.description = assembly.get('@description', '')
         company = assembly.get('@company', '')
-        copyrght = assembly.get('@copyright', '')
-        support_url = assembly.get('@supportInformation', '')
+        package.copyrght = assembly.get('@copyright', '')
+        package.homepage_url = assembly.get('@supportInformation', '')
 
         assembly_identity = assembly.get('assemblyIdentity', {})
-        name = assembly_identity.get('@name', '')
-        version = assembly_identity.get('@version', '')
 
         parties = []
         if company:
@@ -46,14 +54,5 @@ def parse(cls, location):
                     role='owner',
                 )
             )
-
-        yield models.PackageData(
-            datasource_id=cls.datasource_id,
-            type=cls.default_package_type,
-            name=name,
-            version=version,
-            description=description,
-            homepage_url=support_url,
-            parties=parties,
-            copyright=copyrght,
-        )
+        package.parties = parties
+        yield package
diff --git a/src/scancode/api.py b/src/scancode/api.py
index 7d3edbf1516..7c866aeed0c 100644
--- a/src/scancode/api.py
+++ b/src/scancode/api.py
@@ -248,7 +248,13 @@ def get_licenses(
 SCANCODE_DEBUG_PACKAGE_API = os.environ.get('SCANCODE_DEBUG_PACKAGE_API', False)
 
 
-def _get_package_data(location, application=True, system=False, **kwargs):
+def _get_package_data(
+        location,
+        application=True,
+        system=False,
+        purl_only=False,
+        **kwargs
+    ):
     """
     Return a mapping of package manifest information detected in the file at ``location``.
     Include ``application`` packages (such as pypi) and/or ``system`` packages.
@@ -261,7 +267,8 @@ def _get_package_data(location, application=True, system=False, **kwargs):
         return recognize_package_data(
             location=location,
             application=application,
-            system=system
+            system=system,
+            purl_only=purl_only,
         ) or []
 
     except Exception as e:
@@ -291,7 +298,13 @@ def get_package_info(location, **kwargs):
     return dict(packages=[p.to_dict() for p in packages])
 
 
-def get_package_data(location, application=True, system=False, **kwargs):
+def get_package_data(
+        location,
+        application=True,
+        system=False,
+        purl_only=False,
+        **kwargs
+    ):
     """
     Return a mapping of package manifest information detected in the file at
     `location`.
@@ -304,6 +317,7 @@ def get_package_data(location, application=True, system=False, **kwargs):
         location=location,
         application=application,
         system=system,
+        purl_only=purl_only,
         **kwargs,
     ) or []