From 33e30aacc3b69b5d90b32d4d8e16bc9c77c43835 Mon Sep 17 00:00:00 2001 From: Alexey Shamrin Date: Wed, 15 Mar 2006 13:39:11 +0000 Subject: [PATCH] Tutorial draft (not finished) and documentation framework improvements * a first step to a new documentation framework, using Markdown syntax, with some extensions (detailed in txt2myt.py docstrings): * `rel:something` for internal links * `{@name=something}` to override default header names (used when linking) * `{python}` to force code block to use Python syntax highlighting (not needed when using examples with `>>>` prompt) * txt2myt.py -- converter from .txt to .myt * a draft of tutorial.txt, which uses new syntax * testdocs.py -- check examples in documentation using doctest (currently only in tutorial.txt) --- doc/build/content/document_base.myt | 1 + doc/build/content/trailmap.myt | 7 +- doc/build/content/tutorial.txt | 151 +++ doc/build/lib/markdown.py | 1746 +++++++++++++++++++++++++++ doc/build/testdocs.py | 49 + doc/build/txt2myt.py | 250 ++++ 6 files changed, 2201 insertions(+), 3 deletions(-) create mode 100644 doc/build/content/tutorial.txt create mode 100644 doc/build/lib/markdown.py create mode 100644 doc/build/testdocs.py create mode 100644 doc/build/txt2myt.py diff --git a/doc/build/content/document_base.myt b/doc/build/content/document_base.myt index f37857e25f..f45a022e10 100644 --- a/doc/build/content/document_base.myt +++ b/doc/build/content/document_base.myt @@ -3,6 +3,7 @@ <%python scope="global"> files = [ + 'tutorial', 'trailmap', 'pooling', 'dbengine', diff --git a/doc/build/content/trailmap.myt b/doc/build/content/trailmap.myt index d672b07886..e0ea2d1b6c 100644 --- a/doc/build/content/trailmap.myt +++ b/doc/build/content/trailmap.myt @@ -1,7 +1,8 @@ <%flags>inherit='document_base.myt' -<%attr>title='Introduction' -<&|doclib.myt:item, name="intro", description="Introduction" &> -

SQLAlchemy features a lot of tools and patterns to help in every area of writing applications that talk to relational databases. To achieve this, it has a lot of areas of functionality which work together to provide a cohesive package. Ultimately, just a little bit of familiarity with each concept is all thats needed to get off the ground.

+<%attr>title='How to Read this Manual' +<&|doclib.myt:item, name="howtoread", description="How to Read this Manual" &> + +

That said, here's two quick links that summarize the two most prominent features of SQLAlchemy:

+ >>> print i + INSERT INTO users (user_id, user_name, password) VALUES (?, ?, ?) + +Call `execute()` method of the prepared query to actually add users: + + >>> for name in ['Tom', 'Dick', 'Harry']: # doctest:+ELLIPSIS + ... i.execute(user_name = name) + + ... + >>> i.execute(user_name = 'Mary', password = 'secure') # doctest:+ELLIPSIS + + +If possible, SQLAlchemy will bind values into a prepared query using native database API. This allows for better performance, because the database may cache compiled representation of the statement and reuse it between executions with different values. Also, when using bound values, you need not worry about [SQL injection][] attacks. + +[SQL injection]: http://en.wikipedia.org/wiki/SQL_injection + +Main documentation: [sql_insert](rel:sql_insert). + +### Querying the Table + +Let's check that the data we have put into `users` table is actually there. The procedure is analogous to insert example above. The difference is that you have to call `select()` method now: + + >>> s = users.select() + >>> print s + SELECT users.user_id, users.user_name, users.password + FROM users + >>> r = s.execute() + +Now we won't ignore return value of `execute()`: + + >>> r # doctest:+ELLIPSIS + + >>> r.keys + ['user_id', 'user_name', 'password'] + >>> row = r.fetchone() + >>> row['user_name'] + u'Tom' + >>> r.fetchall() + [(2, u'Dick', None), (3, u'Harry', None), (4, u'Mary', u'secure')] + +Main documentation: [sql_select](rel:sql_select). + +### Related Table + +Main documentation: [sql](rel:sql). + +### Fancier Querying {@name=fancyquery} + +Main documentation: [sql](rel:sql). + +### Data Mapping {@name=mapping} + +Main documentation: [datamapping](rel:datamapping), [adv_datamapping](rel:adv_datamapping). + +### Transactions + +Main documentation: [unitofwork](rel:unitofwork), [dbengine_transactions](rel:dbengine_transactions). + +Conclusion +---------- diff --git a/doc/build/lib/markdown.py b/doc/build/lib/markdown.py new file mode 100644 index 0000000000..5531f8dd3a --- /dev/null +++ b/doc/build/lib/markdown.py @@ -0,0 +1,1746 @@ +#!/usr/bin/env python + +""" +==================================================================== +IF YOU ARE LOOKING TO EXTEND MARKDOWN, SEE THE "FOOTNOTES" SECTION +==================================================================== + +Python-Markdown +=============== + +Converts Markdown to HTML. Basic usage as a module: + + import markdown + html = markdown.markdown(your_text_string) + +Started by [Manfred Stienstra](http://www.dwerg.net/). Continued and +maintained by [Yuri Takhteyev](http://www.freewisdom.org). + +Project website: http://www.freewisdom.org/projects/python-markdown +Contact: yuri [at] freewisdom.org + +License: GPL 2 (http://www.gnu.org/copyleft/gpl.html) or BSD + +Version: 1.3 (Feb. 28, 2006) + +For changelog, see end of file +""" + +import re, sys, os, random + +# set debug level: 3 none, 2 critical, 1 informative, 0 all +(VERBOSE, INFO, CRITICAL, NONE) = range(4) + +MESSAGE_THRESHOLD = CRITICAL + +def message(level, text) : + if level >= MESSAGE_THRESHOLD : + print text + + +# --------------- CONSTANTS YOU MIGHT WANT TO MODIFY ----------------- + +# all tabs will be expanded to up to this many spaces +TAB_LENGTH = 4 +ENABLE_ATTRIBUTES = 1 + + +# --------------- CONSTANTS YOU _SHOULD NOT_ HAVE TO CHANGE ---------- + +FN_BACKLINK_TEXT = "zz1337820767766393qq" +# a template for html placeholders +HTML_PLACEHOLDER_PREFIX = "qaodmasdkwaspemas" +HTML_PLACEHOLDER = HTML_PLACEHOLDER_PREFIX + "%dajkqlsmdqpakldnzsdfls" + +BLOCK_LEVEL_ELEMENTS = ['p', 'div', 'blockquote', 'pre', 'table', + 'dl', 'ol', 'ul', 'script', 'noscript', + 'form', 'fieldset', 'iframe', 'math', 'ins', + 'del', 'hr', 'hr/'] + +def is_block_level (tag) : + return ( (tag in BLOCK_LEVEL_ELEMENTS) or + (tag[0] == 'h' and tag[1] in "0123456789") ) + +""" +====================================================================== +========================== NANODOM =================================== +====================================================================== + +The three classes below implement some of the most basic DOM +methods. I use this instead of minidom because I need a simpler +functionality and do not want to require additional libraries. + +Importantly, NanoDom does not do normalization, which is what we +want. It also adds extra white space when converting DOM to string +""" + + +class Document : + + def appendChild(self, child) : + self.documentElement = child + child.parent = self + self.entities = {} + + def createElement(self, tag) : + el = Element(tag) + el.doc = self + return el + + def createTextNode(self, text) : + node = TextNode(text) + node.doc = self + return node + + def createEntityReference(self, entity): + if entity not in self.entities: + self.entities[entity] = EntityReference(entity) + return self.entities[entity] + + def toxml (self) : + return self.documentElement.toxml() + + def normalizeEntities(self, text) : + + pairs = [ #("&", "&"), + ("<", "<"), + (">", ">"), + ("\"", """)] + + for old, new in pairs : + text = text.replace(old, new) + return text + + def find(self, test) : + return self.documentElement.find(test) + + def unlink(self) : + self.documentElement.unlink() + self.documentElement = None + + +class Element : + + def __init__ (self, tag) : + self.type = "element" + self.nodeName = tag + self.attributes = [] + self.attribute_values = {} + self.childNodes = [] + + def unlink(self) : + for child in self.childNodes : + if child.type == "element" : + child.unlink() + self.childNodes = None + + def setAttribute(self, attr, value) : + if not attr in self.attributes : + self.attributes.append(attr) + + self.attribute_values[attr] = value + + def insertChild(self, position, child) : + self.childNodes.insert(position, child) + child.parent = self + + def removeChild(self, child) : + self.childNodes.remove(child) + + def replaceChild(self, oldChild, newChild) : + position = self.childNodes.index(oldChild) + self.removeChild(oldChild) + self.insertChild(position, newChild) + + def appendChild(self, child) : + self.childNodes.append(child) + child.parent = self + + def handleAttributes(self) : + pass + + def find(self, test, depth=0) : + """ Returns a list of descendants that pass the test function """ + matched_nodes = [] + for child in self.childNodes : + if test(child) : + matched_nodes.append(child) + if child.type == "element" : + matched_nodes += child.find(test, depth+1) + return matched_nodes + + def toxml(self): + if ENABLE_ATTRIBUTES : + for child in self.childNodes: + child.handleAttributes() + buffer = "" + if self.nodeName in ['h1', 'h2', 'h3', 'h4'] : + buffer += "\n" + elif self.nodeName in ['li'] : + buffer += "\n " + buffer += "<" + self.nodeName + for attr in self.attributes : + value = self.attribute_values[attr] + value = self.doc.normalizeEntities(value) + buffer += ' %s="%s"' % (attr, value) + if self.childNodes : + buffer += ">" + for child in self.childNodes : + buffer += child.toxml() + if self.nodeName == 'p' : + buffer += "\n" + elif self.nodeName == 'li' : + buffer += "\n " + buffer += "" % self.nodeName + else : + buffer += "/>" + if self.nodeName in ['p', 'li', 'ul', 'ol', + 'h1', 'h2', 'h3', 'h4'] : + buffer += "\n" + return buffer + + +class TextNode : + + def __init__ (self, text) : + self.type = "text" + self.value = text + self.attrRegExp = re.compile(r'\{@([^\}]*)=([^\}]*)}') # {@id=123} + + def attributeCallback(self, match) : + self.parent.setAttribute(match.group(1), match.group(2)) + + def handleAttributes(self) : + self.value = self.attrRegExp.sub(self.attributeCallback, self.value) + + def toxml(self) : + text = self.value + if not text.startswith(HTML_PLACEHOLDER_PREFIX): + if self.parent.nodeName == "p" : + text = text.replace("\n", "\n ") + elif (self.parent.nodeName == "li" + and self.parent.childNodes[0]==self): + text = "\n " + text.replace("\n", "\n ") + text = self.doc.normalizeEntities(text) + return text + + +class EntityReference: + + def __init__(self, entity): + self.type = "entity_ref" + self.entity = entity + + def handleAttributes(self): + pass + + def toxml(self): + return "&" + self.entity + ";" + + +""" +====================================================================== +========================== PRE-PROCESSORS ============================ +====================================================================== + +Preprocessors munge source text before we start doing anything too +complicated. + +Each preprocessor implements a "run" method that takes a pointer to +a list of lines of the document, modifies it as necessary and +returns either the same pointer or a pointer to a new list. +""" + +class HeaderPreprocessor : + + """ + Replaces underlined headers with hashed headers to avoid + the nead for lookahead later. + """ + + def run (self, lines) : + + for i in range(len(lines)) : + if not lines[i] : + continue + + if (i+1 <= len(lines) + and lines[i+1] + and lines[i+1][0] in ['-', '=']) : + + underline = lines[i+1].strip() + + if underline == "="*len(underline) : + lines[i] = "# " + lines[i].strip() + lines[i+1] = "" + elif underline == "-"*len(underline) : + lines[i] = "## " + lines[i].strip() + lines[i+1] = "" + + return lines + +HEADER_PREPROCESSOR = HeaderPreprocessor() + +class LinePreprocessor : + """Deals with HR lines (needs to be done before processing lists)""" + + def run (self, lines) : + for i in range(len(lines)) : + if self._isLine(lines[i]) : + lines[i] = "

" + return lines + + def _isLine(self, block) : + """Determines if a block should be replaced with an

around it - append the paragraph bits directly + # onto parent_elem + el = parent_elem + else : + # Otherwise make a "p" element + el = self.doc.createElement("p") + parent_elem.appendChild(el) + + for item in list : + el.appendChild(item) + + if theRest : + theRest = theRest[1:] # skip the first (blank) line + + self._processSection(parent_elem, theRest, inList) + + + + def _processUList(self, parent_elem, lines, inList) : + self._processList(parent_elem, lines, inList, + listexpr='ul', tag = 'ul') + + def _processOList(self, parent_elem, lines, inList) : + self._processList(parent_elem, lines, inList, + listexpr='ol', tag = 'ol') + + + def _processList(self, parent_elem, lines, inList, listexpr, tag) : + """Given a list of document lines starting with a list item, + finds the end of the list, breaks it up, and recursively + processes each list item and the remainder of the text file. + + @param parent_elem: A dom element to which the content will be added + @param lines: a list of lines + @param inList: a level + @returns: None""" + + ul = self.doc.createElement(tag) # ul might actually be '

' + parent_elem.appendChild(ul) + + looseList = 0 + + # Make a list of list items + items = [] + item = -1 + + i = 0 # a counter to keep track of where we are + + for line in lines : + + loose = 0 + if not line.strip() : + # If we see a blank line, this _might_ be the end of the list + i += 1 + loose = 1 + + # Find the next non-blank line + for j in range(i, len(lines)) : + if lines[j].strip() : + next = lines[j] + break + else : + # There is no more text => end of the list + break + + # Check if the next non-blank line is still a part of the list + if ( RE.regExp[listexpr].match(next) or + RE.regExp['tabbed'].match(next) ): + # get rid of any white space in the line + items[item].append(line.strip()) + looseList = loose or looseList + continue + else : + break # found end of the list + + # Now we need to detect list items (at the current level) + # while also detabing child elements if necessary + + for expr in [listexpr, 'tabbed']: + + m = RE.regExp[expr].match(line) + if m : + if expr == listexpr : # We are looking at a new item + if m.group(1) : + items.append([m.group(1)]) + item += 1 + elif expr == 'tabbed' : # This line needs to be detabbed + items[item].append(m.group(4)) #after the 'tab' + + i += 1 + break + else : + items[item].append(line) # Just regular continuation + i += 1 # added on 2006.02.25 + else : + i += 1 + + # Add the dom elements + for item in items : + li = self.doc.createElement("li") + ul.appendChild(li) + + self._processSection(li, item, inList + 1, looseList = looseList) + + # Process the remaining part of the section + + self._processSection(parent_elem, lines[i:], inList) + + + def _linesUntil(self, lines, condition) : + """ A utility function to break a list of lines upon the + first line that satisfied a condition. The condition + argument should be a predicate function. + """ + + i = -1 + for line in lines : + i += 1 + if condition(line) : break + else : + i += 1 + return lines[:i], lines[i:] + + def _processQuote(self, parent_elem, lines, inList) : + """Given a list of document lines starting with a quote finds + the end of the quote, unindents it and recursively + processes the body of the quote and the remainder of the + text file. + + @param parent_elem: DOM element to which the content will be added + @param lines: a list of lines + @param inList: a level + @returns: None """ + + dequoted = [] + i = 0 + for line in lines : + m = RE.regExp['quoted'].match(line) + if m : + dequoted.append(m.group(1)) + i += 1 + else : + break + else : + i += 1 + + blockquote = self.doc.createElement('blockquote') + parent_elem.appendChild(blockquote) + + self._processSection(blockquote, dequoted, inList) + self._processSection(parent_elem, lines[i:], inList) + + + + + def _processCodeBlock(self, parent_elem, lines, inList) : + """Given a list of document lines starting with a code block + finds the end of the block, puts it into the dom verbatim + wrapped in ("

") and recursively processes the
+           the remainder of the text file.
+
+           @param parent_elem: DOM element to which the content will be added
+           @param lines: a list of lines
+           @param inList: a level
+           @returns: None"""
+
+        detabbed, theRest = self.blockGuru.detectTabbed(lines)
+
+        pre = self.doc.createElement('pre')
+        code = self.doc.createElement('code')
+        parent_elem.appendChild(pre)
+        pre.appendChild(code)
+        text = "\n".join(detabbed).rstrip()+"\n"
+        text = text.replace("&", "&")
+        code.appendChild(self.doc.createTextNode(text))
+        self._processSection(parent_elem, theRest, inList)
+
+
+    def _handleInline(self,  line):
+        """Transform a Markdown line with inline elements to an XHTML fragment.
+
+        Note that this function works recursively: we look for a
+        pattern, which usually splits the paragraph in half, and then
+        call this function on the two parts.
+
+        This function uses auxiliary objects called inline patterns.
+        See notes on inline patterns above.
+
+        @param item: A block of Markdown text
+        @return: A list of NanoDomnodes """
+        if not(line):
+            return [self.doc.createTextNode(' ')]
+        # two spaces at the end of the line denote a 

+        #if line.endswith('  '):
+        #    list = self._handleInline( line.rstrip())
+        #    list.append(self.doc.createElement('br'))
+        #    return list
+        #
+        # ::TODO:: Replace with a preprocessor
+
+        for pattern in self.inlinePatterns :
+            list = self._applyPattern( line, pattern)
+            if list: return list
+
+        return [self.doc.createTextNode(line)]
+
+    def _applyPattern(self,  line, pattern) :
+        """ Given a pattern name, this function checks if the line
+            fits the pattern, creates the necessary elements and
+            recursively calls _handleInline (via. _inlineRecurse)
+
+        @param line: the text to be processed
+        @param pattern: the pattern to be checked
+
+        @returns: the appropriate newly created NanoDom element if the
+                  pattern matches, None otherwise.
+        """
+
+        # match the line to pattern's pre-compiled reg exp.
+        # if no match, move on.
+
+        m = pattern.getCompiledRegExp().match(line)
+        if not m :
+            return None
+
+        # if we got a match let the pattern make us a NanoDom node
+        # if it doesn't, move on
+        node = pattern.handleMatch(m, self.doc)
+        if not node :
+            return None
+
+        # determine what we've got to the left and to the right
+
+        left = m.group(1)      # the first match group
+        left_list = self._handleInline(left)
+        right = m.groups()[-1] # the last match group
+        right_list = self._handleInline(right)
+
+        # put the three parts together
+        left_list.append(node)
+        left_list.extend(right_list)
+
+        return left_list
+
+
+    def __str__(self):
+        """Return the document in XHTML format.
+
+        @returns: A serialized XHTML body."""
+        #try :
+        doc = self._transform()
+        xml = doc.toxml()
+        #finally:
+        #    doc.unlink()
+
+        # Let's stick in all the raw html pieces
+
+        for i in range(self.htmlStash.html_counter) :
+            xml = xml.replace("%s\n" % (HTML_PLACEHOLDER % i),
+                              self.htmlStash.rawHtmlBlocks[i] + "\n")
+            xml = xml.replace(HTML_PLACEHOLDER % i,
+                              self.htmlStash.rawHtmlBlocks[i])
+
+        xml = xml.replace(FN_BACKLINK_TEXT, "↩")
+
+        # And return everything but the top level tag
+        xml = xml.strip()[23:-7]
+
+        return xml
+
+
+    toString = __str__
+
+
+"""
+========================= FOOTNOTES =================================
+
+This section adds footnote handling to markdown.  It can be used as
+an example for extending python-markdown with relatively complex
+functionality.  While in this case the extension is included inside
+the module itself, it could just as easily be added from outside the
+module.  Not that all markdown classes above are ignorant about
+footnotes.  All footnote functionality is provided separately and
+then added to the markdown instance at the run time.
+
+Footnote functionality is attached by calling extendMarkdown()
+method of FootnoteExtension.  The method also registers the
+extension to allow it's state to be reset by a call to reset()
+method.
+"""
+
+class FootnoteExtension :
+
+    def __init__ (self) :
+        self.DEF_RE = re.compile(r'(\ ?\ ?\ ?)\[\^([^\]]*)\]:\s*(.*)')
+        self.SHORT_USE_RE = re.compile(r'\[\^([^\]]*)\]', re.M) # [^a]
+        self.reset()
+
+    def extendMarkdown(self, md) :
+
+        self.md = md
+
+        # Stateless extensions do not need to be registered
+        md.registerExtension(self)
+
+        # Insert a preprocessor before ReferencePreprocessor
+        index = md.preprocessors.index(REFERENCE_PREPROCESSOR)
+        preprocessor = FootnotePreprocessor(self)
+        preprocessor.md = md
+        md.preprocessors.insert(index, preprocessor)
+
+        # Insert an inline pattern before ImageReferencePattern
+        FOOTNOTE_RE = r'\[\^([^\]]*)\]' # blah blah [^1] blah
+        index = md.inlinePatterns.index(IMAGE_REFERENCE_PATTERN)
+        md.inlinePatterns.insert(index, FootnotePattern(FOOTNOTE_RE, self))
+
+        # Insert a post-processor that would actually add the footnote div
+        md.postprocessors.append(FootnotePostprocessor(self))
+
+    def reset(self) :
+        # May be called by Markdown is state reset is desired
+
+        self.footnote_suffix = "-" + str(int(random.random()*1000000000))
+        self.used_footnotes={}
+        self.footnotes = {}
+
+    def setFootnote(self, id, text) :
+        self.footnotes[id] = text
+
+    def makeFootnoteId(self, num) :
+        return 'fn%d%s' % (num, self.footnote_suffix)
+
+    def makeFootnoteRefId(self, num) :
+        return 'fnr%d%s' % (num, self.footnote_suffix)
+
+    def makeFootnotesDiv (self, doc) :
+        """Creates the div with class='footnote' and populates it with
+           the text of the footnotes.
+
+           @returns: the footnote div as a dom element """
+
+        if not self.footnotes.keys() :
+            return None
+
+        div = doc.createElement("div")
+        div.setAttribute('class', 'footnote')
+        hr = doc.createElement("hr")
+        div.appendChild(hr)
+        ol = doc.createElement("ol")
+        div.appendChild(ol)
+
+        footnotes = [(self.used_footnotes[id], id)
+                     for id in self.footnotes.keys()]
+        footnotes.sort()
+
+        for i, id in footnotes :
+            li = doc.createElement('li')
+            li.setAttribute('id', self.makeFootnoteId(i))
+
+            self.md._processSection(li, self.footnotes[id].split("\n"))
+
+            #li.appendChild(doc.createTextNode(self.footnotes[id]))
+
+            backlink = doc.createElement('a')
+            backlink.setAttribute('href', '#' + self.makeFootnoteRefId(i))
+            backlink.setAttribute('class', 'footnoteBackLink')
+            backlink.setAttribute('title',
+                                  'Jump back to footnote %d in the text' % 1)
+            backlink.appendChild(doc.createTextNode(FN_BACKLINK_TEXT))
+
+            if li.childNodes :
+                node = li.childNodes[-1]
+                if node.type == "text" :
+                    node = li
+                node.appendChild(backlink)
+
+            ol.appendChild(li)
+
+        return div
+
+
+class FootnotePreprocessor :
+
+    def __init__ (self, footnotes) :
+        self.footnotes = footnotes
+
+    def run(self, lines) :
+
+        self.blockGuru = BlockGuru()
+        lines = self._handleFootnoteDefinitions (lines)
+
+        # Make a hash of all footnote marks in the text so that we
+        # know in what order they are supposed to appear.  (This
+        # function call doesn't really substitute anything - it's just
+        # a way to get a callback for each occurence.
+
+        text = "\n".join(lines)
+        self.footnotes.SHORT_USE_RE.sub(self.recordFootnoteUse, text)
+
+        return text.split("\n")
+
+
+    def recordFootnoteUse(self, match) :
+
+        id = match.group(1)
+        id = id.strip()
+        nextNum = len(self.footnotes.used_footnotes.keys()) + 1
+        self.footnotes.used_footnotes[id] = nextNum
+
+
+    def _handleFootnoteDefinitions(self, lines) :
+        """Recursively finds all footnote definitions in the lines.
+
+            @param lines: a list of lines of text
+            @returns: a string representing the text with footnote
+                      definitions removed """
+
+        i, id, footnote = self._findFootnoteDefinition(lines)
+
+        if id :
+
+            plain = lines[:i]
+
+            detabbed, theRest = self.blockGuru.detectTabbed(lines[i+1:])
+
+            self.footnotes.setFootnote(id,
+                                       footnote + "\n"
+                                       + "\n".join(detabbed))
+
+            more_plain = self._handleFootnoteDefinitions(theRest)
+            return plain + [""] + more_plain
+
+        else :
+            return lines
+
+    def _findFootnoteDefinition(self, lines) :
+        """Finds the first line of a footnote definition.
+
+            @param lines: a list of lines of text
+            @returns: the index of the line containing a footnote definition """
+
+        counter = 0
+        for line in lines :
+            m = self.footnotes.DEF_RE.match(line)
+            if m :
+                return counter, m.group(2), m.group(3)
+            counter += 1
+        return counter, None, None
+
+
+class FootnotePattern (BasePattern) :
+
+    def __init__ (self, pattern, footnotes) :
+
+        BasePattern.__init__(self, pattern)
+        self.footnotes = footnotes
+
+    def handleMatch(self, m, doc) :
+        sup = doc.createElement('sup')
+        a = doc.createElement('a')
+        sup.appendChild(a)
+        id = m.group(2)
+        num = self.footnotes.used_footnotes[id]
+        sup.setAttribute('id', self.footnotes.makeFootnoteRefId(num))
+        a.setAttribute('href', '#' + self.footnotes.makeFootnoteId(num))
+        a.appendChild(doc.createTextNode(str(num)))
+        return sup
+
+class FootnotePostprocessor :
+
+    def __init__ (self, footnotes) :
+        self.footnotes = footnotes
+
+    def run(self, doc) :
+        footnotesDiv = self.footnotes.makeFootnotesDiv(doc)
+        if footnotesDiv :
+            doc.documentElement.appendChild(footnotesDiv)
+
+# ====================================================================
+
+def markdown(text) :
+    message(VERBOSE, "in markdown.py, received text:\n%s" % text)
+    return str(Markdown(text))
+
+def markdownWithFootnotes(text):
+    message(VERBOSE, "Running markdown with footnotes, "
+            + "received text:\n%s" % text)
+    md = Markdown()
+    footnoteExtension = FootnoteExtension()
+    footnoteExtension.extendMarkdown(md)
+    md.source = text
+
+    return str(md)
+
+def test_markdown(args):
+    """test markdown at the command line.
+        in each test, arg 0 is the module name"""
+    print "\nTEST 1: no arguments on command line"
+    cmd_line(["markdown.py"])
+    print "\nTEST 2a: 1 argument on command line: a good option"
+    cmd_line(["markdown.py","-footnotes"])
+    print "\nTEST 2b: 1 argument on command line: a bad option"
+    cmd_line(["markdown.py","-foodnotes"])
+    print "\nTEST 3: 1 argument on command line: non-existent input file"
+    cmd_line(["markdown.py","junk.txt"])
+    print "\nTEST 4: 1 argument on command line: existing input file"
+    lines = """
+Markdown text with[^1]:
+
+2. **bold text**,
+3. *italic text*.
+
+Then more:
+
+    beginning of code block;
+    another line of code block.
+    
+    a second paragraph of code block.
+
+more text to end our file.
+
+[^1]: "italic" means emphasis.
+"""
+    fid = "markdown-test.txt"
+    f1 = open(fid, 'w+')
+    f1.write(lines)
+    f1.close()
+    cmd_line(["markdown.py",fid])
+    print "\nTEST 5: 2 arguments on command line: nofootnotes and input file"
+    cmd_line(["markdown.py","-nofootnotes", fid])
+    print "\nTEST 6: 2 arguments on command line: footnotes and input file"
+    cmd_line(["markdown.py","-footnotes", fid])
+    print "\nTEST 7: 3 arguments on command line: nofootnotes,inputfile, outputfile"
+    fidout = "markdown-test.html"
+    cmd_line(["markdown.py","-nofootnotes", fid, fidout])
+
+
+def get_vars(args):
+    """process the command-line args received; return usable variables"""
+    #firstly get the variables
+
+    message(VERBOSE, "in get_vars(), args: %s" % args) 
+
+    if len(args) <= 1:
+        option, inFile, outFile = (None, None, None)
+    elif len(args) >= 4:
+        option, inFile, outFile = args[1:4]
+    elif len(args) == 3:
+        temp1, temp2 = args[1:3]
+        if temp1[0] == '-':
+            #then we have an option and inFile
+            option, inFile, outFile = temp1, temp2, None
+        else:
+            #we have no option, so we must have inFile and outFile
+            option, inFile, outFile = None, temp1, temp2
+    else:
+        #len(args) = 2
+        #we have only one usable arg: might be an option or a file
+        temp1 = args[1]
+        
+        message(VERBOSE, "our single arg is: %s" % str(temp1))
+
+        if temp1[0] == '-':
+            #then we have an option 
+            option, inFile, outFile = temp1, None, None
+        else:
+            #we have no option, so we must have inFile
+            option, inFile, outFile = None, temp1, None
+    
+    message(VERBOSE,
+            "prior to validation, option: %s, inFile: %s, outFile: %s" %
+            (str(option), str(inFile), str(outFile),))
+    
+    return option, inFile, outFile
+
+
+USAGE = """
+\nUsing markdown.py:
+
+    python markdown.py [option] input_file_with_markdown.txt [output_file.html]
+
+Options:
+
+    -footnotes or -fn   : generate markdown with footnotes
+    -test or -t         : run a self-test
+    -help or -h         : print this message
+
+"""
+    
+VALID_OPTIONS = ['footnotes','nofootnotes', 'fn', 'test', 't', 'f',
+                 'help', 'h']
+
+EXPANDED_OPTIONS =  { "fn" : "footnotes",
+                      "t"  : "test",
+                      "h"  : "help" }
+
+
+def validate_option(option) :
+
+    """ Check if the option makes sense and print an appropriate message
+        if it isn't.
+        
+        @return: valid option string or None
+    """
+
+    #now validate the variables
+    if (option is not None):
+        if (len(option) > 1 and option[1:] in VALID_OPTIONS) :
+            option = option[1:]
+
+            if option in EXPANDED_OPTIONS.keys() :
+                option = EXPANDED_OPTIONS[option]
+            return option
+        else:
+            message(CRITICAL,
+                    "\nSorry, I don't understand option %s" % option)
+            message(CRITICAL, USAGE)
+            return None
+
+
+def validate_input_file(inFile) :        
+    """ Check if the input file is specified and exists.
+
+        @return: valid input file path or None
+    """
+
+    if not inFile :
+        message(CRITICAL,
+                "\nI need an input filename.\n")
+        message(CRITICAL, USAGE)
+        return None
+    
+        
+    if os.access(inFile, os.R_OK):
+        return inFile
+    else :
+        message(CRITICAL, "Sorry, I can't find input file %s" % str(inFile))
+        return None
+
+    
+            
+
+def cmd_line(args):
+
+    message(VERBOSE, "in cmd_line with args: %s" % args)
+
+    option, inFile, outFile = get_vars(args)
+
+    if option :
+        option = validate_option(option)
+        if not option : return
+
+    if option == "help" :
+        message(CRITICAL, USAGE)
+        return
+    elif option == "test" :
+        test_markdown(None)
+        return
+
+    inFile = validate_input_file(inFile)
+    if not inFile :
+        return
+    else :
+        input = file(inFile).read()
+
+    message(VERBOSE, "Validated command line parameters:" +             
+             "\n\toption: %s, \n\tinFile: %s, \n\toutFile: %s" % (
+             str(option), str(inFile), str(outFile),))
+
+    if option == "footnotes" :
+        md_function = markdownWithFootnotes
+    else :
+        md_function = markdown
+
+    if outFile is None:
+        print md_function(input)
+    else:
+        output = md_function(input)
+        f1 = open(outFile, "w+")
+        f1.write(output)
+        f1.close()
+        
+        if os.access(outFile, os.F_OK):
+            message(INFO, "Successfully wrote %s" % outFile)
+        else:
+            message(INFO, "Failed to write %s" % outFile)
+
+
+if __name__ == '__main__':
+    """ Run Markdown from the command line.
+        Set debug = 3 at top of file to get diagnostic output"""
+    args = sys.argv
+        
+    #set testing=1 to test the command-line response of markdown.py
+    testing = 0
+    if testing:
+        test_markdown(args)
+    else:
+        cmd_line(args)
+
+"""
+CHANGELOG
+=========
+
+Feb. 28, 2006: Clean-up and command-line handling by Stewart
+Midwinter. (Version 1.3)
+
+Feb. 24, 2006: Fixed a bug with the last line of the list appearing
+again as a separate paragraph.  Incorporated Chris Clark's "mailto"
+patch.  Added support for 
 at the end of lines ending in two or
+more spaces.  Fixed a crashing bug when using ImageReferencePattern.
+Added several utility methods to Nanodom.  (Version 1.2)
+
+Jan. 31, 2006: Added "hr" and "hr/" to BLOCK_LEVEL_ELEMENTS and
+changed 
 to .  (Thanks to Sergej Chodarev.)
+
+Nov. 26, 2005: Fixed a bug with certain tabbed lines inside lists
+getting wrapped in .  (v. 1.1)
+
+Nov. 19, 2005: Made ">>\s*[a-zA-Z_]\w*\s*=\s*create_engine\('sqlite',\s*\{'filename':\s*')" + oldfile+ "('\}\)$)"
+    engine = re.compile(engine, re.MULTILINE)
+    s, n = re.subn(engine, r'\1' + newfile + r'\2', s, 1)
+    if not n:
+        raise ValueError("Couldn't find suitable create_engine call to replace '%s' in it" % oldfile)
+    return s
+
+filename = 'content/tutorial.txt'
+s = open(filename).read()
+s = replace_file(s, 'tutorial.db', ':memory:')
+teststring(s, filename)
+
diff --git a/doc/build/txt2myt.py b/doc/build/txt2myt.py
new file mode 100644
index 0000000000..789a0e39a7
--- /dev/null
+++ b/doc/build/txt2myt.py
@@ -0,0 +1,250 @@
+"""txt2myt.py -- translate all .txt files in a `content` directory to
+.myt template, assuming .txt conform to Markdown syntax.
+"""
+
+import sys
+import string
+import re
+
+import elementtree.ElementTree as et
+
+sys.path.insert(0, './lib')
+import markdown
+
+class MyElementTree(et.ElementTree):
+    def _write(self, file, node, encoding, namespaces):
+        """With support for myghty 'tags'"""
+        if node.tag is MyghtyTag:
+            params = ', '.join(['%s="%s"' % i for i in node.items()])
+            pipe = ''
+            if node.text or len(node):
+                pipe = '|'
+            comma = ''
+            if params:
+                comma = ', '
+            file.write('<&%s%s%s%s&>' % (pipe, node.name, comma, params))
+            if pipe:
+                if node.text:
+                    file.write(node.text)
+                for n in node:
+                    self._write(file, n, encoding, namespaces)
+                file.write("")
+            if node.tail:
+                file.write(node.tail)
+        else:
+            et.ElementTree._write(self, file, node, encoding, namespaces)
+
+# The same as et.dump, but using MyElementTree
+def dump(elem):
+    # debugging
+    if not isinstance(elem, et.ElementTree):
+        elem = MyElementTree(elem)
+    elem.write(sys.stdout)
+    tail = elem.getroot().tail
+    if not tail or tail[-1] != "\n":
+        sys.stdout.write("\n")
+
+# The same as et.tostring, but using MyElementTree
+def tostring(element, encoding=None):
+    class dummy:
+        pass
+    data = []
+    file = dummy()
+    file.write = data.append
+    MyElementTree(element).write(file, encoding)
+    return string.join(data, "")
+
+def MyghtyTag(name_, attrib_={}, **extra):
+    """Can be used with ElementTree in places where Element is required"""
+    element = et.Element(MyghtyTag, attrib_, **extra)
+    element.name = name_
+    return element
+
+CODE_BLOCK = 'formatting.myt:code'
+CODE_BLOCK_MARKER = '{python}'
+DOCTEST_DIRECTIVES = re.compile(r'#\s*doctest:\s*[+-]\w+(,[+-]\w+)*\s*$', re.M)
+LINK = 'formatting.myt:link'
+LINK_MARKER = 'rel:'
+SECTION = 'doclib.myt:item'
+
+def process_code_blocks(tree):
+    """Replace ... with Myghty tags, if it contains:
+    * '{python}'
+    or
+    * '>>> '
+
+    Note: '{python}' will be removed
+    Note: also remove all doctest directives
+    """
+    parent = get_parent_map(tree)
+    keyword = CODE_BLOCK_MARKER
+
+    def replace_pre_with_myt(pre, text):
+        text = re.sub(DOCTEST_DIRECTIVES, '', text)
+        pre_parent = parent[pre]
+        tag = MyghtyTag(CODE_BLOCK)
+        tag.text = text
+        tag.tail = pre.tail
+        pre_parent[index(pre_parent, pre)] = tag
+
+    for precode in tree.findall('.//pre/code'):
+        if precode.text.lstrip().startswith(keyword):
+            text = precode.text.lstrip()
+            text = text.replace(keyword, '', 1).lstrip()
+            replace_pre_with_myt(parent[precode], text)
+        elif precode.text.lstrip().startswith('>>> '):
+            replace_pre_with_myt(parent[precode], precode.text)
+
+def process_rel_href(tree):
+    """Replace all YYY with Myghty tags
+
+    If XXX is the same as YYY, text attribute for Myghty tag will not be
+    provided.
+    """
+    parent = get_parent_map(tree)
+    for a in tree.findall('.//a'):
+        if a.get('href').startswith(LINK_MARKER):
+            text = a.text
+            path = a.get('href')[len(LINK_MARKER):]
+            if text == path:
+                tag = MyghtyTag(LINK, path=path)
+            else:
+                tag = MyghtyTag(LINK, path=path, text=text)
+            tag.tail = a.tail
+            a_parent = parent[a]
+            a_parent[index(a_parent, a)] = tag
+
+def process_headers(tree):
+    """Replace all , 
... with Mighty tags
+    """
+    title = [None]
+    def process(tree):
+        while True:
+            i = find_header_index(tree)
+            if i is None:
+                return
+            node = tree[i]
+            level = int(node.tag[1])
+            start, end = i, end_of_header(tree, level, i+1)
+            content = tree[start+1:end]
+            description = node.text.strip()
+            if title[0] is None:
+                title[0] = description
+            name = node.get('name')
+            if name is None:
+                name = description.split()[0].lower()
+
+            tag = MyghtyTag(SECTION, name=name, description=description)
+            tag.text = node.tail + '\n'
+            tag.tail = '\n'
+            tag[:] = content
+            tree[start:end] = [tag]
+
+            process(tag)
+
+    process(tree)
+    return title[0]
+
+def index(parent, item):
+    for n, i in enumerate(parent):
+        if i is item:
+            return n
+
+def find_header_index(tree):
+    for i, node in enumerate(tree):
+        if is_header(node):
+            return i
+
+def is_header(node):
+    t = node.tag
+    return (isinstance(t, str) and len(t) == 2 and t[0] == 'h' 
+            and t[1] in '123456789')
+
+def end_of_header(tree, level, start):
+    for i, node in enumerate(tree[start:]):
+        if is_header(node) and int(node.tag[1]) <= level:
+            return start + i
+    return len(tree)
+
+def get_parent_map(tree):
+    return dict((c, p) for p in tree.getiterator() for c in p)
+ 
+def html2myghtydoc(html):
+    """Convert HTML to Myghty template (for SQLAlchemy's doc framework)
+
+    html -- string containing HTML document, which:
+    * should be without surrounding  or  tags
+    * should have only one top-level header (usually 
). The text of the
+      header will be used as a title of Myghty template.
+
+    Things converter treats specially:
+    * YYY will result in:
+      <&formatting.myt:link, path="XXX", text="YYY"&>
+
+      If XXX is the same as YYY, text attribute will be absent. E.g.
+      XXX will result in:
+      <&formatting.myt:link, path="XXX"&>
+
+    * All header tags (
, 
...) will be replaced. E.g. 
+
+      
Great title
Section content goes here
+
+      will turn into:
+
+      <&|doclib.myt:item, name="great", description="Great title"&>
+      Section content goes here
+      
+
+      Note that by default the value of `name` parameter is a lower-case version
+      of the first word of the title. If you want to override it, supply
+      `name` attribute to header tag, e.g.:
+      Great title...
+      will turn into
+      <&|doclib.myt:item, name="title", description="Great title"&>...
+
+      (Note that to give this attribute in Markdown, you can
+      write: {@name=title}.)
+
+    * If you have {python} marker inside , which is inside , it will
+      replaced with Myghty tag. E.g.:
+
+      {python} print 'hello, world!'
+
+      will turn into
+
+      <&|formatting.myt:code&>print 'hello, world!'
+
+      You don't need to write {python} marker if you use examples starting with
+      Python prompt: >>>
+
+      If you use doctest directives, they will be removed from output.
+
+      (Note that  and  is what Markdown outputs for pre-formatted
+      code blocks.)
+    """
+
+    tree = et.fromstring("" + html + "")
+
+    process_code_blocks(tree)
+    process_rel_href(tree)
+    title = process_headers(tree)
+
+    header = "<%flags>inherit='document_base.myt'\n"
+    header += "<%attr>title='" + title + "'\n"
+    header += "\n"
+
+    # discard surrounding  tag
+    body = ''.join(tostring(e) for e in tree[:])
+
+    return header + body
+
+if __name__ == '__main__':
+    import glob
+    for inname in glob.glob('content/*.txt'):
+        outname = inname[:-3] + 'myt'
+        print inname, '->', outname
+        input = file(inname).read()
+        html = markdown.markdown(input)
+        myt = html2myghtydoc(html)
+        file(outname, 'w').write(myt)
-- 
2.47.2

,

,

...) will be replaced. E.g. + +

Great title

Great title