Addition of a glob function

This is a *bit* of a hack. Add missing documentation for `read_bytes()` 
function.

docs/builtins.md

@@ -382,6 +382,11 @@ The name `c` will have the value `"o w"`.
 
 ## File functions
 
+`glob(` *pattern* `)`
+: Return an *n*-vector of strings representing file paths matching the given
+shell "glob" pattern. Files are matched relative to the directory containing
+the running program.
+
 `csv(` *filename* `,` *row* `)`
 : Return a vector of values obtained by reading a
 specific *row* (indexed from *0*) from the CSV file with the given *filename*;
@@ -391,3 +396,9 @@ the row into numeric values.
 `read(` *filename* `)`
 : Returns a single string value containing the entire text of *filename* (this
 function intelligently caches).
+
+`read_bytes(` *filename* `)`
+: Returns a numeric vector value containing the entire text of *filename* (this
+function intelligently caches) as a sequence of bytes. This is a *very*
+memory-inefficient function as each byte will be represented by a
+double-precision floating point value.

src/flitter/cache.py

@@ -29,6 +29,10 @@ def __init__(self, path, absolute):
     def suffix(self):
         return self._path.suffix
 
+    @property
+    def parent(self):
+        return self._path.parent
+
     def exists(self):
         return self._path.exists()
 

src/flitter/language/functions.pyx

@@ -162,6 +162,11 @@ cdef class normal(uniform):
         return self.R * cos(self.th)
 
 
+@context_func
+def glob(Context context, Vector pattern):
+    return Vector._coerce(sorted(str(p.resolve()) for p in context.path.parent.glob(pattern.as_string())))
+
+
 @context_func
 def read_text(Context context, Vector filename):
     cdef str path = filename.as_string()
@@ -178,6 +183,15 @@ def read_bytes(Context context, Vector filename):
     return null_
 
 
+@context_func
+def read_csv(Context context, Vector filename, Vector row_number):
+    cdef str path = str(filename)
+    row = row_number.match(1, int)
+    if filename and row is not None:
+        return SharedCache.get_with_root(path, context.path).read_csv_vector(row)
+    return null_
+
+
 def ordv(Vector text):
     cdef str string = text.as_string()
     return Vector._coerce([ord(ch) for ch in string])
@@ -202,15 +216,6 @@ def split(Vector text):
     return Vector._coerce(text.as_string().rstrip('\n').split('\n'))
 
 
-@context_func
-def read_csv(Context context, Vector filename, Vector row_number):
-    cdef str path = str(filename)
-    row = row_number.match(1, int)
-    if filename and row is not None:
-        return SharedCache.get_with_root(path, context.path).read_csv_vector(row)
-    return null_
-
-
 def lenv(Vector xs not None):
     cdef Vector ys = Vector.__new__(Vector)
     ys.allocate_numbers(1)
@@ -1025,6 +1030,7 @@ STATIC_FUNCTIONS = {
 DYNAMIC_FUNCTIONS = {
     'debug': Vector(debug),
     'sample': Vector(sample),
+    'glob': Vector(glob),
     'csv': Vector(read_csv),
     'read': Vector(read_text),
     'read_bytes': Vector(read_bytes),