#!/usr/bin/env python #Copyright ReportLab Europe Ltd. 2000-2004 #see license.txt for license details #history http://www.reportlab.co.uk/cgi-bin/viewcvs.cgi/public/reportlab/trunk/reportlab/tools/docco/docpy.py """Generate documentation from live Python objects. This is an evolving module that allows to generate documentation for python modules in an automated fashion. The idea is to take live Python objects and inspect them in order to use as much mean- ingful information as possible to write in some formatted way into different types of documents. In principle a skeleton captures the gathered information and makes it available via a certain API to formatters that use it in whatever way they like to produce something of interest. The API allows for adding behaviour in subclasses of these formatters, such that, e.g. for certain classes it is possible to trigger special actions like displaying a sample image of a class that represents some graphical widget, say. Type the following for usage info: python docpy.py -h """ # Much inspired by Ka-Ping Yee's htmldoc.py. # Needs the inspect module. # Dinu Gherman __version__ = '0.8' import sys, os, re, types, string, getopt, copy, time from string import find, join, split, replace, expandtabs, rstrip from reportlab.pdfgen import canvas from reportlab.lib import colors from reportlab.lib.units import inch, cm from reportlab.lib.pagesizes import A4 from reportlab.lib import enums from reportlab.lib.enums import TA_CENTER, TA_LEFT from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle from reportlab.platypus.flowables import Flowable, Spacer from reportlab.platypus.paragraph import Paragraph from reportlab.platypus.flowables \ import Flowable, Preformatted,Spacer, Image, KeepTogether, PageBreak from reportlab.platypus.tableofcontents import TableOfContents from reportlab.platypus.xpreformatted import XPreformatted from reportlab.platypus.frames import Frame from reportlab.platypus.doctemplate \ import PageTemplate, BaseDocTemplate from reportlab.platypus.tables import TableStyle, Table import inspect #################################################################### # # Stuff needed for building PDF docs. # #################################################################### def mainPageFrame(canvas, doc): "The page frame used for all PDF documents." canvas.saveState() pageNumber = canvas.getPageNumber() canvas.line(2*cm, A4[1]-2*cm, A4[0]-2*cm, A4[1]-2*cm) canvas.line(2*cm, 2*cm, A4[0]-2*cm, 2*cm) if pageNumber > 1: canvas.setFont('Times-Roman', 12) canvas.drawString(4 * inch, cm, "%d" % pageNumber) if hasattr(canvas, 'headerLine'): # hackish headerline = string.join(canvas.headerLine, ' \215 ') # bullet canvas.drawString(2*cm, A4[1]-1.75*cm, headerline) canvas.setFont('Times-Roman', 8) msg = "Generated with reportlab.lib.docpy. See http://www.reportlab.com!" canvas.drawString(2*cm, 1.65*cm, msg) canvas.restoreState() class MyTemplate(BaseDocTemplate): "The document template used for all PDF documents." _invalidInitArgs = ('pageTemplates',) def __init__(self, filename, **kw): frame1 = Frame(2.5*cm, 2.5*cm, 15*cm, 25*cm, id='F1') self.allowSplitting = 0 apply(BaseDocTemplate.__init__, (self, filename), kw) self.addPageTemplates(PageTemplate('normal', [frame1], mainPageFrame)) def afterFlowable(self, flowable): "Takes care of header line, TOC and outline entries." if flowable.__class__.__name__ == 'Paragraph': f = flowable name7 = f.style.name[:7] name8 = f.style.name[:8] # Build a list of heading parts. # So far, this is the *last* item on the *previous* page... if name7 == 'Heading' and not hasattr(self.canv, 'headerLine'): self.canv.headerLine = [] if name8 == 'Heading0': self.canv.headerLine = [f.text] # hackish elif name8 == 'Heading1': if len(self.canv.headerLine) == 2: del self.canv.headerLine[-1] elif len(self.canv.headerLine) == 3: del self.canv.headerLine[-1] del self.canv.headerLine[-1] self.canv.headerLine.append(f.text) elif name8 == 'Heading2': if len(self.canv.headerLine) == 3: del self.canv.headerLine[-1] self.canv.headerLine.append(f.text) if name7 == 'Heading': # Register TOC entries. headLevel = int(f.style.name[7:]) self.notify('TOCEntry', (headLevel, flowable.getPlainText(), self.page)) # Add PDF outline entries. c = self.canv title = f.text key = str(hash(f)) try: if headLevel == 0: isClosed = 0 else: isClosed = 1 c.bookmarkPage(key) c.addOutlineEntry(title, key, level=headLevel, closed=isClosed) except ValueError: pass #################################################################### # # Utility functions (Ka-Ping Yee). # #################################################################### def htmlescape(text): "Escape special HTML characters, namely &, <, >." return replace(replace(replace(text, '&', '&'), '<', '<'), '>', '>') def htmlrepr(object): return htmlescape(repr(object)) def defaultformat(object): return '=' + htmlrepr(object) def getdoc(object): result = inspect.getdoc(object) if not result: try: result = inspect.getcomments(object) except: pass return result and rstrip(result) + '\n' or '' def reduceDocStringLength(docStr): "Return first line of a multiline string." return split(docStr, '\n')[0] #################################################################### # # More utility functions # #################################################################### def makeHtmlSection(text, bgcolor='#FFA0FF'): """Create HTML code for a section. This is usually a header for all classes or functions. u """ text = htmlescape(expandtabs(text)) result = [] result.append("""""") result.append("""
""" % bgcolor) result.append("""

%s

""" % text) result.append("""
""") result.append('') return join(result, '\n') def makeHtmlSubSection(text, bgcolor='#AAA0FF'): """Create HTML code for a subsection. This is usually a class or function name. """ text = htmlescape(expandtabs(text)) result = [] result.append("""""") result.append("""
""" % bgcolor) result.append("""

%s

""" % text) result.append("""
""") result.append('') return join(result, '\n') def makeHtmlInlineImage(text): """Create HTML code for an inline image. """ return """%s""" % (text, text) #################################################################### # # Core "standard" docpy classes # #################################################################### class PackageSkeleton0: """A class collecting 'interesting' information about a package.""" pass # Not yet! class ModuleSkeleton0: """A class collecting 'interesting' information about a module.""" def __init__(self): # This is an ad-hoc, somewhat questionable 'data structure', # but for the time being it serves its purpose and is fairly # self-contained. self.module = {} self.functions = {} self.classes = {} # Might need more like this, later. def getModuleName(self): """Return the name of the module being treated.""" return self.module['name'] # These inspect methods all rely on the inspect module. def inspect(self, object): """Collect information about a given object.""" self.moduleSpace = object # Very non-OO, left for later... if inspect.ismodule(object): self._inspectModule(object) elif inspect.isclass(object): self._inspectClass(object) elif inspect.ismethod(object): self._inspectMethod(object) elif inspect.isfunction(object): self._inspectFunction(object) elif inspect.isbuiltin(object): self._inspectBuiltin(object) else: msg = "Don't know how to document this kind of object." raise TypeError, msg def _inspectModule(self, object): """Collect information about a given module object.""" name = object.__name__ self.module['name'] = name if hasattr(object, '__version__'): self.module['version'] = object.__version__ cadr = lambda list: list[1] modules = map(cadr, inspect.getmembers(object, inspect.ismodule)) classes, cdict = [], {} for key, value in inspect.getmembers(object, inspect.isclass): if (inspect.getmodule(value) or object) is object: classes.append(value) cdict[key] = cdict[value] = '#' + key functions, fdict = [], {} for key, value in inspect.getmembers(object, inspect.isroutine): #if inspect.isbuiltin(value) or inspect.getmodule(value) is object: functions.append(value) fdict[key] = '#-' + key if inspect.isfunction(value): fdict[value] = fdict[key] for c in classes: for base in c.__bases__: key, modname = base.__name__, base.__module__ if modname != name and sys.modules.has_key(modname): module = sys.modules[modname] if hasattr(module, key) and getattr(module, key) is base: if not cdict.has_key(key): cdict[key] = cdict[base] = modname + '.txt#' + key ## doc = getdoc(object) or 'No doc string.' doc = getdoc(object) self.module['doc'] = doc if modules: self.module['importedModules'] = map(lambda m:m.__name__, modules) if classes: for item in classes: self._inspectClass(item, fdict, cdict) if functions: for item in functions: self._inspectFunction(item, fdict, cdict) def _inspectClass(self, object, functions={}, classes={}): """Collect information about a given class object.""" name = object.__name__ bases = object.__bases__ results = [] if bases: parents = [] for base in bases: parents.append(base) self.classes[name] = {} if bases: self.classes[name]['bases'] = parents methods, mdict = [], {} for key, value in inspect.getmembers(object, inspect.ismethod): methods.append(value) mdict[key] = mdict[value] = '#' + name + '-' + key if methods: if not self.classes[name].has_key('methods'): self.classes[name]['methods'] = {} for item in methods: self._inspectMethod(item, functions, classes, mdict, name) ## doc = getdoc(object) or 'No doc string.' doc = getdoc(object) self.classes[name]['doc'] = doc def _inspectMethod(self, object, functions={}, classes={}, methods={}, clname=''): """Collect information about a given method object.""" self._inspectFunction(object.im_func, functions, classes, methods, clname) def _inspectFunction(self, object, functions={}, classes={}, methods={}, clname=''): """Collect information about a given function object.""" try: args, varargs, varkw, defaults = inspect.getargspec(object) argspec = inspect.formatargspec( args, varargs, varkw, defaults, defaultformat=defaultformat) except TypeError: argspec = '( ... )' ## doc = getdoc(object) or 'No doc string.' doc = getdoc(object) if object.__name__ == '': decl = [' lambda ', argspec[1:-1]] # print ' %s' % decl # Do something with lambda functions as well... # ... else: decl = object.__name__ if not clname: self.functions[object.__name__] = {'signature':argspec, 'doc':doc} else: theMethods = self.classes[clname]['methods'] if not theMethods.has_key(object.__name__): theMethods[object.__name__] = {} theMethod = theMethods[object.__name__] theMethod['signature'] = argspec theMethod['doc'] = doc def _inspectBuiltin(self, object): """Collect information about a given built-in.""" print object.__name__ + '( ... )' def walk(self, formatter): """Call event methods in a visiting formatter.""" s = self f = formatter # The order is fixed, but could be made flexible # with one more template method... # Module modName = s.module['name'] modDoc = s.module['doc'] imported = s.module.get('importedModules', []) imported.sort() # f.indentLevel = f.indentLevel + 1 f.beginModule(modName, modDoc, imported) # Classes f.indentLevel = f.indentLevel + 1 f.beginClasses(s.classes.keys()) items = s.classes.items() items.sort() for k, v in items: cDoc = s.classes[k]['doc'] bases = s.classes[k].get('bases', []) f.indentLevel = f.indentLevel + 1 f.beginClass(k, cDoc, bases) # This if should move out of this method. if not s.classes[k].has_key('methods'): s.classes[k]['methods'] = {} # Methods #f.indentLevel = f.indentLevel + 1 f.beginMethods(s.classes[k]['methods'].keys()) items = s.classes[k]['methods'].items() items.sort() for m, v in items: mDoc = v['doc'] sig = v['signature'] f.indentLevel = f.indentLevel + 1 f.beginMethod(m, mDoc, sig) f.indentLevel = f.indentLevel - 1 f.endMethod(m, mDoc, sig) #f.indentLevel = f.indentLevel - 1 f.endMethods(s.classes[k]['methods'].keys()) f.indentLevel = f.indentLevel - 1 f.endClass(k, cDoc, bases) # And what about attributes?! f.indentLevel = f.indentLevel - 1 f.endClasses(s.classes.keys()) # Functions f.indentLevel = f.indentLevel + 1 f.beginFunctions(s.functions.keys()) items = s.functions.items() items.sort() for k, v in items: doc = v['doc'] sig = v['signature'] f.indentLevel = f.indentLevel + 1 f.beginFunction(k, doc, sig) f.indentLevel = f.indentLevel - 1 f.endFunction(k, doc, sig) f.indentLevel = f.indentLevel - 1 f.endFunctions(s.functions.keys()) #f.indentLevel = f.indentLevel - 1 f.endModule(modName, modDoc, imported) # Constants?! #################################################################### # # Core "standard" docpy document builders # #################################################################### class DocBuilder0: """An abstract class to document the skeleton of a Python module. Instances take a skeleton instance s and call their s.walk() method. The skeleton, in turn, will walk over its tree structure while generating events and calling equivalent methods from a specific interface (begin/end methods). """ fileSuffix = None def __init__(self, skeleton=None): self.skeleton = skeleton self.packageName = None self.indentLevel = 0 def write(self, skeleton=None): if skeleton: self.skeleton = skeleton self.skeleton.walk(self) # Event-method API, called by associated skeleton instances. # In fact, these should raise a NotImplementedError, but for now we # just don't do anything here. # The following four methods are *not* called by skeletons! def begin(self, name='', typ=''): pass def end(self): pass # Methods for packaging should move into a future PackageSkeleton... def beginPackage(self, name): self.packageName = name def endPackage(self, name): pass # Only this subset is really called by associated skeleton instances. def beginModule(self, name, doc, imported): pass def endModule(self, name, doc, imported): pass def beginClasses(self, names): pass def endClasses(self, names): pass def beginClass(self, name, doc, bases): pass def endClass(self, name, doc, bases): pass def beginMethods(self, names): pass def endMethods(self, names): pass def beginMethod(self, name, doc, sig): pass def endMethod(self, name, doc, sig): pass def beginFunctions(self, names): pass def endFunctions(self, names): pass def beginFunction(self, name, doc, sig): pass def endFunction(self, name, doc, sig): pass class AsciiDocBuilder0(DocBuilder0): """Document the skeleton of a Python module in ASCII format. The output will be an ASCII file with nested lines representing the hiearchical module structure. Currently, no doc strings are listed. """ fileSuffix = '.txt' outLines = [] indentLabel = ' ' def end(self): # This if should move into DocBuilder0... if self.packageName: self.outPath = self.packageName + self.fileSuffix elif self.skeleton: self.outPath = self.skeleton.getModuleName() + self.fileSuffix else: self.outPath = '' if self.outPath: file = open(self.outPath, 'w') for line in self.outLines: file.write(line + '\n') file.close() def beginPackage(self, name): DocBuilder0.beginPackage(self, name) lev, label = self.indentLevel, self.indentLabel self.outLines.append('%sPackage: %s' % (lev*label, name)) self.outLines.append('') def beginModule(self, name, doc, imported): append = self.outLines.append lev, label = self.indentLevel, self.indentLabel self.outLines.append('%sModule: %s' % (lev*label, name)) ## self.outLines.append('%s%s' % ((lev+1)*label, reduceDocStringLength(doc))) append('') if imported: self.outLines.append('%sImported' % ((lev+1)*label)) append('') for m in imported: self.outLines.append('%s%s' % ((lev+2)*label, m)) append('') def beginClasses(self, names): if names: lev, label = self.indentLevel, self.indentLabel self.outLines.append('%sClasses' % (lev*label)) self.outLines.append('') def beginClass(self, name, doc, bases): append = self.outLines.append lev, label = self.indentLevel, self.indentLabel if bases: bases = map(lambda b:b.__name__, bases) # hack append('%s%s(%s)' % (lev*label, name, join(bases, ', '))) else: append('%s%s' % (lev*label, name)) return ## append('%s%s' % ((lev+1)*label, reduceDocStringLength(doc))) self.outLines.append('') def endClass(self, name, doc, bases): self.outLines.append('') def beginMethod(self, name, doc, sig): append = self.outLines.append lev, label = self.indentLevel, self.indentLabel append('%s%s%s' % (lev*label, name, sig)) ## append('%s%s' % ((lev+1)*label, reduceDocStringLength(doc))) ## append('') def beginFunctions(self, names): if names: lev, label = self.indentLevel, self.indentLabel self.outLines.append('%sFunctions' % (lev*label)) self.outLines.append('') def endFunctions(self, names): self.outLines.append('') def beginFunction(self, name, doc, sig): append = self.outLines.append lev, label = self.indentLevel, self.indentLabel self.outLines.append('%s%s%s' % (lev*label, name, sig)) ## append('%s%s' % ((lev+1)*label, reduceDocStringLength(doc))) ## append('') class HtmlDocBuilder0(DocBuilder0): "A class to write the skeleton of a Python source in HTML format." fileSuffix = '.html' outLines = [] def begin(self, name='', typ=''): self.outLines.append("""""") self.outLines.append("""""") def end(self): if self.packageName: self.outPath = self.packageName + self.fileSuffix elif self.skeleton: self.outPath = self.skeleton.getModuleName() + self.fileSuffix else: self.outPath = '' if self.outPath: file = open(self.outPath, 'w') self.outLines.append('') for line in self.outLines: file.write(line + '\n') file.close() def beginPackage(self, name): DocBuilder0.beginPackage(self, name) self.outLines.append("""%s""" % name) self.outLines.append("""""") self.outLines.append("""

%s

""" % name) self.outLines.append('') def beginModule(self, name, doc, imported): if not self.packageName: self.outLines.append("""%s""" % name) self.outLines.append("""""") self.outLines.append("""

%s

""" % name) self.outLines.append('') for line in split(doc, '\n'): self.outLines.append("""%s""" % htmlescape(line)) self.outLines.append('
') self.outLines.append('') if imported: self.outLines.append(makeHtmlSection('Imported Modules')) self.outLines.append("""""") def beginClasses(self, names): self.outLines.append(makeHtmlSection('Classes')) def beginClass(self, name, doc, bases): DocBuilder0.beginClass(self, name, doc, bases) ## # Keep an eye on the base classes. ## self.currentBaseClasses = bases if bases: bases = map(lambda b:b.__name__, bases) # hack self.outLines.append(makeHtmlSubSection('%s(%s)' % (name, join(bases, ', ')))) else: self.outLines.append(makeHtmlSubSection('%s' % name)) for line in split(doc, '\n'): self.outLines.append("""%s""" % htmlescape(line)) self.outLines.append('
') self.outLines.append('') def beginMethods(self, names): pass ## if names: ## self.outLines.append('

Method Interface

') ## self.outLines.append('') def beginMethod(self, name, doc, sig): self.beginFunction(name, doc, sig) def beginFunctions(self, names): self.outLines.append(makeHtmlSection('Functions')) def beginFunction(self, name, doc, sig): append = self.outLines.append append("""
%s%s
""" % (name, sig)) append('') for line in split(doc, '\n'): append("""
%s
""" % htmlescape(line)) append('
') append('
') append('') class PdfDocBuilder0(DocBuilder0): "Document the skeleton of a Python module in PDF format." fileSuffix = '.pdf' def makeHeadingStyle(self, level, typ=None, doc=''): "Make a heading style for different types of module content." if typ in ('package', 'module', 'class'): style = ParagraphStyle(name='Heading'+str(level), fontName = 'Courier-Bold', fontSize=14, leading=18, spaceBefore=12, spaceAfter=6) elif typ in ('method', 'function'): if doc: style = ParagraphStyle(name='Heading'+str(level), fontName = 'Courier-Bold', fontSize=12, leading=18, firstLineIndent=-18, leftIndent=36, spaceBefore=0, spaceAfter=-3) else: style = ParagraphStyle(name='Heading'+str(level), fontName = 'Courier-Bold', fontSize=12, leading=18, firstLineIndent=-18, leftIndent=36, spaceBefore=0, spaceAfter=0) else: style = ParagraphStyle(name='Heading'+str(level), fontName = 'Times-Bold', fontSize=14, leading=18, spaceBefore=12, spaceAfter=6) return style def begin(self, name='', typ=''): styleSheet = getSampleStyleSheet() self.code = styleSheet['Code'] self.bt = styleSheet['BodyText'] self.story = [] # Cover page t = time.gmtime(time.time()) timeString = time.strftime("%Y-%m-%d %H:%M", t) self.story.append(Paragraph('Documentation for %s "%s"' % (typ, name), self.bt)) self.story.append(Paragraph('Generated by: docpy.py version %s' % __version__, self.bt)) self.story.append(Paragraph('Date generated: %s' % timeString, self.bt)) self.story.append(Paragraph('Format: PDF', self.bt)) self.story.append(PageBreak()) # Table of contents toc = TableOfContents() self.story.append(toc) self.story.append(PageBreak()) def end(self): if self.outPath is not None: pass elif self.packageName: self.outPath = self.packageName + self.fileSuffix elif self.skeleton: self.outPath = self.skeleton.getModuleName() + self.fileSuffix else: self.outPath = '' print 'output path is %s' % self.outPath if self.outPath: doc = MyTemplate(self.outPath) doc.multiBuild(self.story) def beginPackage(self, name): DocBuilder0.beginPackage(self, name) story = self.story story.append(Paragraph(name, self.makeHeadingStyle(self.indentLevel, 'package'))) def beginModule(self, name, doc, imported): story = self.story bt = self.bt story.append(Paragraph(name, self.makeHeadingStyle(self.indentLevel, 'module'))) if doc: story.append(XPreformatted(htmlescape(doc), bt)) story.append(XPreformatted('', bt)) if imported: story.append(Paragraph('Imported modules', self.makeHeadingStyle(self.indentLevel + 1))) for m in imported: p = Paragraph('\201 %s' % m, bt) p.style.bulletIndent = 10 p.style.leftIndent = 18 story.append(p) def endModule(self, name, doc, imported): DocBuilder0.endModule(self, name, doc, imported) self.story.append(PageBreak()) def beginClasses(self, names): self.story.append(Paragraph('Classes', self.makeHeadingStyle(self.indentLevel))) def beginClass(self, name, doc, bases): bt = self.bt story = self.story if bases: bases = map(lambda b:b.__name__, bases) # hack story.append(Paragraph('%s(%s)' % (name, join(bases, ', ')), self.makeHeadingStyle(self.indentLevel, 'class'))) else: story.append(Paragraph(name, self.makeHeadingStyle(self.indentLevel, 'class'))) if doc: story.append(XPreformatted(htmlescape(doc), bt)) story.append(XPreformatted('', bt)) def beginMethod(self, name, doc, sig): bt = self.bt story = self.story story.append(Paragraph(name+sig, self.makeHeadingStyle(self.indentLevel, 'method', doc))) if doc: story.append(XPreformatted(htmlescape(doc), bt)) story.append(XPreformatted('', bt)) def beginFunctions(self, names): if names: self.story.append(Paragraph('Functions', self.makeHeadingStyle(self.indentLevel))) def beginFunction(self, name, doc, sig): bt = self.bt story = self.story story.append(Paragraph(name+sig, self.makeHeadingStyle(self.indentLevel, 'function'))) if doc: story.append(XPreformatted(htmlescape(doc), bt)) story.append(XPreformatted('', bt)) class UmlPdfDocBuilder0(PdfDocBuilder0): "Document the skeleton of a Python module with UML class diagrams." fileSuffix = '.pdf' def begin(self, name='', typ=''): styleSheet = getSampleStyleSheet() self.h1 = styleSheet['Heading1'] self.h2 = styleSheet['Heading2'] self.h3 = styleSheet['Heading3'] self.code = styleSheet['Code'] self.bt = styleSheet['BodyText'] self.story = [] self.classCompartment = '' self.methodCompartment = [] def beginModule(self, name, doc, imported): story = self.story h1, h2, h3, bt = self.h1, self.h2, self.h3, self.bt styleSheet = getSampleStyleSheet() bt1 = styleSheet['BodyText'] story.append(Paragraph(name, h1)) story.append(XPreformatted(doc, bt1)) if imported: story.append(Paragraph('Imported modules', self.makeHeadingStyle(self.indentLevel + 1))) for m in imported: p = Paragraph('\201 %s' % m, bt1) p.style.bulletIndent = 10 p.style.leftIndent = 18 story.append(p) def endModule(self, name, doc, imported): self.story.append(PageBreak()) PdfDocBuilder0.endModule(self, name, doc, imported) def beginClasses(self, names): h1, h2, h3, bt = self.h1, self.h2, self.h3, self.bt if names: self.story.append(Paragraph('Classes', h2)) def beginClass(self, name, doc, bases): self.classCompartment = '' self.methodCompartment = [] if bases: bases = map(lambda b:b.__name__, bases) # hack self.classCompartment = '%s(%s)' % (name, join(bases, ', ')) else: self.classCompartment = name def endClass(self, name, doc, bases): h1, h2, h3, bt, code = self.h1, self.h2, self.h3, self.bt, self.code styleSheet = getSampleStyleSheet() bt1 = styleSheet['BodyText'] story = self.story # Use only the first line of the class' doc string -- # no matter how long! (Do the same later for methods) classDoc = reduceDocStringLength(doc) tsa = tableStyleAttributes = [] # Make table with class and method rows # and add it to the story. p = Paragraph('%s' % self.classCompartment, bt) p.style.alignment = TA_CENTER rows = [(p,)] # No doc strings, now... # rows = rows + [(Paragraph('%s' % classDoc, bt1),)] lenRows = len(rows) tsa.append(('BOX', (0,0), (-1,lenRows-1), 0.25, colors.black)) for name, doc, sig in self.methodCompartment: nameAndSig = Paragraph('%s%s' % (name, sig), bt1) rows.append((nameAndSig,)) # No doc strings, now... # docStr = Paragraph('%s' % reduceDocStringLength(doc), bt1) # rows.append((docStr,)) tsa.append(('BOX', (0,lenRows), (-1,-1), 0.25, colors.black)) t = Table(rows, (12*cm,)) tableStyle = TableStyle(tableStyleAttributes) t.setStyle(tableStyle) self.story.append(t) self.story.append(Spacer(1*cm, 1*cm)) def beginMethod(self, name, doc, sig): self.methodCompartment.append((name, doc, sig)) def beginFunctions(self, names): h1, h2, h3, bt = self.h1, self.h2, self.h3, self.bt if names: self.story.append(Paragraph('Functions', h2)) self.classCompartment = chr(171) + ' Module-Level Functions ' + chr(187) self.methodCompartment = [] def beginFunction(self, name, doc, sig): self.methodCompartment.append((name, doc, sig)) def endFunctions(self, names): h1, h2, h3, bt, code = self.h1, self.h2, self.h3, self.bt, self.code styleSheet = getSampleStyleSheet() bt1 = styleSheet['BodyText'] story = self.story if not names: return tsa = tableStyleAttributes = [] # Make table with class and method rows # and add it to the story. p = Paragraph('%s' % self.classCompartment, bt) p.style.alignment = TA_CENTER rows = [(p,)] lenRows = len(rows) tsa.append(('BOX', (0,0), (-1,lenRows-1), 0.25, colors.black)) for name, doc, sig in self.methodCompartment: nameAndSig = Paragraph('%s%s' % (name, sig), bt1) rows.append((nameAndSig,)) # No doc strings, now... # docStr = Paragraph('%s' % reduceDocStringLength(doc), bt1) # rows.append((docStr,)) tsa.append(('BOX', (0,lenRows), (-1,-1), 0.25, colors.black)) t = Table(rows, (12*cm,)) tableStyle = TableStyle(tableStyleAttributes) t.setStyle(tableStyle) self.story.append(t) self.story.append(Spacer(1*cm, 1*cm)) #################################################################### # # Main # #################################################################### def printUsage(): """docpy.py - Automated documentation for Python source code. Usage: python docpy.py [options] [options] -h Print this help message. -f name Use the document builder indicated by 'name', e.g. Ascii, Html, Pdf (default), UmlPdf. -m module Generate document for module named 'module' (default is 'docpy'). 'module' may follow any of these forms: - docpy.py - docpy - c:\\test\\docpy and can be any of these: - standard Python modules - modules in the Python search path - modules in the current directory -p package Generate document for package named 'package'. 'package' may follow any of these forms: - reportlab - reportlab.platypus - c:\\test\\reportlab and can be any of these: - standard Python packages (?) - packages in the Python search path - packages in the current directory -s Silent mode (default is unset). Examples: python docpy.py -h python docpy.py -m docpy.py -f Ascii python docpy.py -m string -f Html python docpy.py -m signsandsymbols.py -f Pdf python docpy.py -p reportlab.platypus -f UmlPdf python docpy.py -p reportlab.lib -s -f UmlPdf """ def documentModule0(pathOrName, builder, opts={}): """Generate documentation for one Python file in some format. This handles Python standard modules like string, custom modules on the Python search path like e.g. docpy as well as modules specified with their full path like C:/tmp/junk.py. The doc file will always be saved in the current directory with a basename equal to that of the module, e.g. docpy. """ cwd = os.getcwd() # Append directory to Python search path if we get one. dirName = os.path.dirname(pathOrName) if dirName: sys.path.append(dirName) # Remove .py extension from module name. if pathOrName[-3:] == '.py': modname = pathOrName[:-3] else: modname = pathOrName # Remove directory paths from module name. if dirName: modname = os.path.basename(modname) # Load the module. try: module = __import__(modname) except: print 'Failed to import %s.' % modname os.chdir(cwd) return # Do the real documentation work. s = ModuleSkeleton0() s.inspect(module) builder.write(s) # Remove appended directory from Python search path if we got one. if dirName: del sys.path[-1] os.chdir(cwd) def _packageWalkCallback((builder, opts), dirPath, files): "A callback function used when waking over a package tree." # Skip __init__ files. files = filter(lambda f:f != '__init__.py', files) files = filter(lambda f:f[-3:] == '.py', files) for f in files: path = os.path.join(dirPath, f) if not opts.get('isSilent', 0): print path builder.indentLevel = builder.indentLevel + 1 documentModule0(path, builder) builder.indentLevel = builder.indentLevel - 1 def documentPackage0(pathOrName, builder, opts={}): """Generate documentation for one Python package in some format. 'pathOrName' can be either a filesystem path leading to a Python package or package name whose path will be resolved by importing the top-level module. The doc file will always be saved in the current directory with a basename equal to that of the package, e.g. reportlab.lib. """ # Did we get a package path with OS-dependant seperators...? if os.sep in pathOrName: path = pathOrName name = os.path.splitext(os.path.basename(path))[0] # ... or rather a package name? else: name = pathOrName package = __import__(name) # Some special care needed for dotted names. if '.' in name: subname = 'package' + name[find(name, '.'):] package = eval(subname) path = os.path.dirname(package.__file__) cwd = os.getcwd() builder.beginPackage(name) os.path.walk(path, _packageWalkCallback, (builder, opts)) builder.endPackage(name) os.chdir(cwd) def main(): "Handle command-line options and trigger corresponding action." opts, args = getopt.getopt(sys.argv[1:], 'hsf:m:p:') # Make an options dictionary that is easier to use. optsDict = {} for k, v in opts: optsDict[k] = v hasOpt = optsDict.has_key # On -h print usage and exit immediately. if hasOpt('-h'): print printUsage.__doc__ sys.exit(0) # On -s set silent mode. isSilent = hasOpt('-s') # On -f set the appropriate DocBuilder to use or a default one. builderClassName = optsDict.get('-f', 'Pdf') + 'DocBuilder0' builder = eval(builderClassName + '()') # Set default module or package to document. if not hasOpt('-p') and not hasOpt('-m'): optsDict['-m'] = 'docpy' # Save a few options for further use. options = {'isSilent':isSilent} # Now call the real documentation functions. if hasOpt('-m'): nameOrPath = optsDict['-m'] if not isSilent: print "Generating documentation for module %s..." % nameOrPath builder.begin(name=nameOrPath, typ='module') documentModule0(nameOrPath, builder, options) elif hasOpt('-p'): nameOrPath = optsDict['-p'] if not isSilent: print "Generating documentation for package %s..." % nameOrPath builder.begin(name=nameOrPath, typ='package') documentPackage0(nameOrPath, builder, options) builder.end() if not isSilent: print "Saved %s." % builder.outPath if __name__ == '__main__': main()