toolslm 0.3.16__tar.gz → 0.3.18__tar.gz

{toolslm-0.3.16/toolslm.egg-info → toolslm-0.3.18}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: toolslm
-Version: 0.3.16
+Version: 0.3.18
 Summary: Tools to make language models a bit easier to use
 Home-page: https://github.com/AnswerDotAI/toolslm
 Author: Jeremy Howard

{toolslm-0.3.16 → toolslm-0.3.18}/settings.ini

@@ -1,7 +1,7 @@
 [DEFAULT]
 repo = toolslm
 lib_name = toolslm
-version = 0.3.16
+version = 0.3.18
 min_python = 3.9
 license = apache2
 black_formatting = False

toolslm-0.3.18/toolslm/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.3.18"

{toolslm-0.3.16 → toolslm-0.3.18}/toolslm/_modidx.py

@@ -44,12 +44,14 @@ d = { 'settings': { 'branch': 'main',
                              'toolslm.xml.files2ctx': ('xml.html#files2ctx', 'toolslm/xml.py'),
                              'toolslm.xml.folder2ctx': ('xml.html#folder2ctx', 'toolslm/xml.py'),
                              'toolslm.xml.folder2ctx_cli': ('xml.html#folder2ctx_cli', 'toolslm/xml.py'),
+                             'toolslm.xml.get_docstring': ('xml.html#get_docstring', 'toolslm/xml.py'),
                              'toolslm.xml.get_mime_text': ('xml.html#get_mime_text', 'toolslm/xml.py'),
                              'toolslm.xml.json_to_xml': ('xml.html#json_to_xml', 'toolslm/xml.py'),
                              'toolslm.xml.mk_doc': ('xml.html#mk_doc', 'toolslm/xml.py'),
                              'toolslm.xml.mk_doctype': ('xml.html#mk_doctype', 'toolslm/xml.py'),
                              'toolslm.xml.nb2xml': ('xml.html#nb2xml', 'toolslm/xml.py'),
                              'toolslm.xml.parse_gh_url': ('xml.html#parse_gh_url', 'toolslm/xml.py'),
+                             'toolslm.xml.py2sigs': ('xml.html#py2sigs', 'toolslm/xml.py'),
                              'toolslm.xml.read_file': ('xml.html#read_file', 'toolslm/xml.py'),
                              'toolslm.xml.repo2ctx': ('xml.html#repo2ctx', 'toolslm/xml.py'),
                              'toolslm.xml.sym2file': ('xml.html#sym2file', 'toolslm/xml.py'),

{toolslm-0.3.16 → toolslm-0.3.18}/toolslm/xml.py

@@ -1,12 +1,12 @@
 # AUTOGENERATED! DO NOT EDIT! File to edit: ../00_xml.ipynb.
 
 # %% auto 0
-__all__ = ['doctype', 'json_to_xml', 'get_mime_text', 'cell2out', 'cell2xml', 'nb2xml', 'mk_doctype', 'mk_doc', 'docs_xml',
-           'read_file', 'files2ctx', 'folder2ctx', 'sym2file', 'sym2folderctx', 'sym2pkgpath', 'sym2pkgctx',
-           'folder2ctx_cli', 'parse_gh_url', 'repo2ctx']
+__all__ = ['doctype', 'json_to_xml', 'get_mime_text', 'cell2out', 'cell2xml', 'nb2xml', 'get_docstring', 'py2sigs', 'mk_doctype',
+           'mk_doc', 'docs_xml', 'read_file', 'files2ctx', 'folder2ctx', 'sym2file', 'sym2folderctx', 'sym2pkgpath',
+           'sym2pkgctx', 'folder2ctx_cli', 'parse_gh_url', 'repo2ctx']
 
 # %% ../00_xml.ipynb
-import hashlib, inspect, xml.etree.ElementTree as ET
+import hashlib, inspect, xml.etree.ElementTree as ET, ast
 from collections import namedtuple
 from ghapi.all import GhApi
 
@@ -50,24 +50,47 @@ def cell2out(o):
     if hasattr(o, 'ename'): return Out(f"{o.ename}: {o.evalue}", type='error')
 
 # %% ../00_xml.ipynb
-def cell2xml(cell, out=True):
+def cell2xml(cell, out=True, ids=True):
     "Convert notebook cell to concise XML format"
     src = ''.join(getattr(cell, 'source', ''))
     f = Code if cell.cell_type=='code' else Md
-    if not out: return f(src)
+    kw = dict(id=cell.id) if ids and hasattr(cell, 'id') else {}
+    if not out: return f(src, **kw)
     parts = [Source(src)]
     out_items = L(getattr(cell,'outputs',[])).map(cell2out).filter()
     if out_items: parts.append(Outs(*out_items))
-    return f(*parts)
+    return f(*parts, **kw)
 
 # %% ../00_xml.ipynb
-def nb2xml(fname=None, nb=None, out=True):
+def nb2xml(fname=None, nb=None, out=True, ids=True):
     "Convert notebook to XML format"
     assert bool(fname)^bool(nb), "Pass either `fname` or `nb`"
     if not nb: nb = dict2obj(fname.read_json())
-    cells_xml = [to_xml(cell2xml(c, out=out), do_escape=False) for c in nb.cells if c.cell_type in ('code','markdown')]
+    cells_xml = [to_xml(cell2xml(c, out=out, ids=ids), do_escape=False) for c in nb.cells if c.cell_type in ('code','markdown')]
     return to_xml(Notebook(*cells_xml), do_escape=False)
 
+# %% ../00_xml.ipynb
+def get_docstring(node, lines):
+    "Get docstring from source lines if present"
+    if not (node.body and isinstance(node.body[0], ast.Expr) and isinstance(node.body[0].value, ast.Constant)): return None
+    doc_node = node.body[0]
+    return '\n'.join(lines[doc_node.lineno-1:doc_node.end_lineno])
+
+def py2sigs(fname=None, src=None):
+    "Return signature+docstring text for all functions and class methods in source"
+    if fname: src = Path(fname).read_text()
+    tree = ast.parse(src)
+    lines = src.splitlines()
+    res = []
+    for node in ast.walk(tree):
+        if isinstance(node, (ast.FunctionDef, ast.AsyncFunctionDef)):
+            body_start = max(node.body[0].lineno - 1, node.lineno)
+            sig = '\n'.join(lines[node.lineno-1:body_start])
+            doc = get_docstring(node, lines)
+            cts = f"{sig}\n{doc}" if doc else sig
+            res.append(cts.strip('\r\n'))
+    return '\n\n'.join(res)
+
 # %% ../00_xml.ipynb
 doctype = namedtuple('doctype', ['src', 'content'])
 
@@ -110,15 +133,16 @@ def docs_xml(docs:list[str],  # The content of each document
     pre = 'Here are some documents for you to reference for your task:\n\n' if prefix else ''
     if srcs is None: srcs = [None]*len(docs)
     if details is None: details = [{}]*len(docs)
-    docs = (mk_doc(i+1, d, s, **kw) for i,(d,s,kw) in enumerate(zip(docs,srcs,details)))
+    docs = (mk_doc(i+1, d, s, **kw) for i,(d,s,kw) in enumerate(zip(docs,srcs,details)) if d.strip())
     kw = dict(title=title) if title else {}
     return pre + to_xml(Documents(*docs, **kw), do_escape=False)
 
 # %% ../00_xml.ipynb
-def read_file(fname, out=True, max_size=None):
+def read_file(fname, out=True, max_size=None, ids=True, sigs_only=False):
     "Read file content, converting notebooks to XML if needed"
     fname = Path(fname)
-    if fname.suffix == '.ipynb': res = nb2xml(fname, out=out)
+    if fname.suffix == '.ipynb': res = nb2xml(fname, out=out, ids=ids)
+    elif fname.suffix == '.py' and sigs_only: res = py2sigs(fname)
     else: res = fname.read_text()
     if max_size and len(res)>max_size: return f"[Skipped: {fname.name} exceeds {max_size} bytes]"
     return res
@@ -130,11 +154,13 @@ def files2ctx(
     out:bool=True, # Include notebook cell outputs?
     srcs:Optional[list]=None, # Use the labels instead of `fnames`
     max_size:int=None, # Skip files larger than this (bytes)
+    ids:bool=True,  # Include cell ids in notebooks?
+    sigs_only:bool=False, # For .py files, only include signatures and docstrings
     **kwargs
 )->str: # XML for LM context
     "Convert files to XML context, handling notebooks"
     fnames = [Path(o) for o in fnames]
-    contents = [read_file(o, out=out, max_size=max_size) for o in fnames]
+    contents = [read_file(o, out=out, max_size=max_size, ids=ids, sigs_only=sigs_only) for o in fnames]
     return docs_xml(contents, srcs or fnames, **kwargs)
 
 # %% ../00_xml.ipynb
@@ -149,6 +175,8 @@ def folder2ctx(
     max_total:int=10_000_000,  # Max total output size in bytes
     readme_first:bool=False,  # Prioritize README files at start of context?
     files_only:bool=False,  # Return dict of {filename: size} instead of context?
+    sigs_only:bool=False,  # Return signatures instead of full text for python files?
+    ids:bool=True,  # Include cell ids in notebooks?
     **kwargs
 )->Union[str,dict]:
     "Convert folder contents to XML context, handling notebooks"
@@ -157,7 +185,7 @@ def folder2ctx(
     if files_only: return {str(f.relative_to(folder)): f.stat().st_size for f in fnames}
     if readme_first: fnames = sorted(fnames, key=lambda f: (0 if 'readme' in f.name.lower() else 1, f))
     srcs = fnames if include_base else [f.relative_to(folder) for f in fnames]
-    res = files2ctx(fnames, prefix=prefix, out=out, srcs=srcs, title=title, max_size=max_size)
+    res = files2ctx(fnames, prefix=prefix, out=out, srcs=srcs, title=title, max_size=max_size, sigs_only=sigs_only, ids=ids)
     suf = f"\n\n[TRUNCATED: output size {{_outsz_}} exceeded max size {max_total} bytes]"
     if max_total and len(res) > max_total: res = truncstr(res, max_total, suf=suf, sizevar='_outsz_')
     return res

{toolslm-0.3.16 → toolslm-0.3.18/toolslm.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: toolslm
-Version: 0.3.16
+Version: 0.3.18
 Summary: Tools to make language models a bit easier to use
 Home-page: https://github.com/AnswerDotAI/toolslm
 Author: Jeremy Howard

toolslm-0.3.16/toolslm/__init__.py

@@ -1 +0,0 @@
-__version__ = "0.3.16"