mirrors/cpython
1
0
Fork 0
mirror of https://github.com/python/cpython.git synced 2025-11-03 11:23:31 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 8

Update the docstring.

Browse source 
Lots of changes to get the paragraph marking to work, and not go into
an infinite recursion.

Start to rationalize markup of method signatures; not complete.

find_all_elements():  similar to getElementsByTagName(), but operates
	on the not-quite-legal fragments we deal with here.
	Simplifies several individual transforms.

fixup_sectionauthors():  Convert \sectionauthor items to <author> just
	after the <section>'s <title>.
This commit is contained in:
Fred Drake 1999-01-28 23:59:58 +00:00
parent 3f3b09657f
commit 7dab6affba
1 changed files with 159 additions and 101 deletions
Download patch file Download diff file Expand all files Collapse all files

232
Doc/tools/sgmlconv/docfixer.py
View file

@ -1,12 +1,7 @@
#! /usr/bin/env python #! /usr/bin/env python
"""Promote the IDs from <label/> elements to the enclosing section / chapter / """Perform massive transformations on a document tree created from the LaTeX
whatever, then remove the <label/> elements. This allows *ML style internal of the Python documentation, and dump the ESIS data for the transformed tree.
linking rather than the bogus LaTeX model.
Note that <label/>s in <title> elements are promoted two steps, since the
<title> elements are artificially created from the section parameter, and the
label really refers to the sectioning construct.
""" """
__version__ = '$Revision$' __version__ = '$Revision$'
@ -26,6 +21,13 @@ class ConversionError(Exception):
DEBUG_PARA_FIXER = 0 DEBUG_PARA_FIXER = 0
if DEBUG_PARA_FIXER:
def para_msg(s):
sys.stderr.write("*** %s\n" % s)
else:
def para_msg(s):
pass
# Workaround to deal with invalid documents (multiple root elements). This # Workaround to deal with invalid documents (multiple root elements). This
# does not indicate a bug in the DOM implementation. # does not indicate a bug in the DOM implementation.
@ -62,6 +64,19 @@ def extract_first_element(doc, gi):
return node return node
def find_all_elements(doc, gi):
nodes = []
if doc.nodeType == xml.dom.core.ELEMENT and doc.tagName == gi:
nodes.append(doc)
for child in doc.childNodes:
if child.nodeType == xml.dom.core.ELEMENT:
if child.tagName == gi:
nodes.append(child)
for node in child.getElementsByTagName(gi):
nodes.append(node)
return nodes
def simplify(doc): def simplify(doc):
# Try to rationalize the document a bit, since these things are simply # Try to rationalize the document a bit, since these things are simply
# not valid SGML/XML documents as they stand, and need a little work. # not valid SGML/XML documents as they stand, and need a little work.
@ -108,31 +123,50 @@ def cleanup_root_text(doc):
doc.removeChild(node) doc.removeChild(node)
def rewrite_desc_entries(doc, argname_gi): def handle_args(doc):
argnodes = doc.getElementsByTagName(argname_gi) for node in find_all_elements(doc, "args"):
for node in argnodes:
parent = node.parentNode parent = node.parentNode
nodes = [] nodes = []
for n in parent.childNodes: for n in parent.childNodes:
if n.nodeType != xml.dom.core.ELEMENT or n.tagName != argname_gi: if n.nodeType != xml.dom.core.ELEMENT or n.tagName != "args":
nodes.append(n) nodes.append(n)
signature = doc.createElement("signature")
signature.appendChild(doc.createTextNode("\n "))
name = doc.createElement("name")
name.appendChild(doc.createTextNode(parent.getAttribute("name")))
parent.removeAttribute("name")
signature.appendChild(name)
desc = doc.createElement("description") desc = doc.createElement("description")
for n in nodes: for n in nodes:
parent.removeChild(n) parent.removeChild(n)
desc.appendChild(n) desc.appendChild(n)
desc.appendChild(doc.createTextNode("\n "))
parent.replaceChild(signature, node)
parent.insertBefore(doc.createTextNode("\n "), signature)
if node.childNodes: if node.childNodes:
# keep the <args>...</args>, newline & indent # keep the <args>...</args>, newline & indent
parent.insertBefore(doc.createText("\n "), node) signature.appendChild(doc.createTextNode("\n "))
else: signature.appendChild(node)
# no arguments, remove the <args/> node
parent.removeChild(node)
parent.appendChild(doc.createText("\n ")) parent.appendChild(doc.createText("\n "))
parent.appendChild(desc) parent.appendChild(desc)
parent.appendChild(doc.createText("\n")) parent.appendChild(doc.createText("\n"))
signature.appendChild(doc.createTextNode("\n "))
def handle_args(doc):
rewrite_desc_entries(doc, "args") def methodline_to_signature(doc, methodline):
rewrite_desc_entries(doc, "constructor-args") signature = doc.createElement("signature")
signature.appendChild(doc.createTextNode("\n "))
name = doc.createElement("name")
name.appendChild(doc.createTextNode(methodline.getAttribute("name")))
signature.appendChild(name)
methodline.parentNode.removeChild(methodline)
if len(methodline.childNodes):
methodline._node.name = "args"
methodline.removeAttribute("name")
signature.appendChild(doc.createTextNode("\n "))
signature.appendChild(methodline)
signature.appendChild(doc.createTextNode("\n "))
return signature
def handle_appendix(doc): def handle_appendix(doc):
@ -165,10 +199,7 @@ def handle_appendix(doc):
def handle_labels(doc): def handle_labels(doc):
for node in doc.childNodes: for label in find_all_elements(doc, "label"):
if node.nodeType == xml.dom.core.ELEMENT:
labels = node.getElementsByTagName("label")
for label in labels:
id = label.getAttribute("id") id = label.getAttribute("id")
if not id: if not id:
continue continue
@ -306,7 +337,8 @@ def create_module_info(doc, section):
if first_data.data[:4] == " ---": if first_data.data[:4] == " ---":
first_data.data = string.lstrip(first_data.data[4:]) first_data.data = string.lstrip(first_data.data[4:])
title._node.name = "short-synopsis" title._node.name = "short-synopsis"
if children[-1].data[-1:] == ".": if children[-1].nodeType == xml.dom.core.TEXT \
and children[-1].data[-1:] == ".":
children[-1].data = children[-1].data[:-1] children[-1].data = children[-1].data[:-1]
section.removeChild(title) section.removeChild(title)
section.removeChild(section.childNodes[0]) section.removeChild(section.childNodes[0])
@ -335,9 +367,7 @@ def create_module_info(doc, section):
def cleanup_synopses(doc): def cleanup_synopses(doc):
for node in doc.childNodes: for node in find_all_elements(doc, "section"):
if node.nodeType == xml.dom.core.ELEMENT \
and node.tagName == "section":
create_module_info(doc, node) create_module_info(doc, node)
@ -361,12 +391,10 @@ def remap_element_names(root, name_map):
def fixup_table_structures(doc): def fixup_table_structures(doc):
# must be done after remap_element_names(), or the tables won't be found # must be done after remap_element_names(), or the tables won't be found
for child in doc.childNodes: for table in find_all_elements(doc, "table"):
if child.nodeType == xml.dom.core.ELEMENT:
tables = child.getElementsByTagName("table")
for table in tables:
fixup_table(doc, table) fixup_table(doc, table)
def fixup_table(doc, table): def fixup_table(doc, table):
# create the table head # create the table head
thead = doc.createElement("thead") thead = doc.createElement("thead")
@ -443,66 +471,81 @@ def move_elements_by_name(doc, source, dest, name, sep=None):
FIXUP_PARA_ELEMENTS = ( FIXUP_PARA_ELEMENTS = (
"chapter", "chapter",
"section", "subsection", "subsubsection", "section", "subsection", "subsubsection",
"paragraph", "subparagraph", "description", "paragraph", "subparagraph",
"opcodedesc", "classdesc", "excdesc", "datadesc",
"funcdesc", "methoddesc", "excdesc", "datadesc", "excdescni", "datadescni",
"funcdescni", "methoddescni", "excdescni", "datadescni", )
RECURSE_INTO_PARA_CONTAINERS = (
"chapter",
"section", "subsection", "subsubsection",
"paragraph", "subparagraph",
"abstract",
"memberdesc", "memberdescni", "datadesc", "datadescni",
) )
PARA_LEVEL_ELEMENTS = ( PARA_LEVEL_ELEMENTS = (
"moduleinfo", "title", "verbatim", "moduleinfo", "title", "verbatim",
"opcodedesc", "classdesc", "opcodedesc", "classdesc",
"funcdesc", "methoddesc", "excdesc", "datadesc", "funcdesc", "methoddesc", "excdesc",
"funcdescni", "methoddescni", "excdescni", "datadescni", "funcdescni", "methoddescni", "excdescni",
"tableii", "tableiii", "tableiv", "localmoduletable", "tableii", "tableiii", "tableiv", "localmoduletable",
"sectionauthor", "sectionauthor", "seealso",
# include <para>, so we can just do it again to get subsequent paras: # include <para>, so we can just do it again to get subsequent paras:
"para", "para",
) )
PARA_LEVEL_PRECEEDERS = ( PARA_LEVEL_PRECEEDERS = (
"index", "indexii", "indexiii", "indexiv", "index", "indexii", "indexiii", "indexiv",
"stindex", "obindex", "COMMENT", "label", "stindex", "obindex", "COMMENT", "label", "input",
"memberline", "memberlineni",
"methodline", "methodlineni",
) )
def fixup_paras(doc): def fixup_paras(doc):
for child in doc.childNodes: for child in doc.childNodes:
if child.nodeType == xml.dom.core.ELEMENT \ if child.nodeType == xml.dom.core.ELEMENT \
and child.tagName in FIXUP_PARA_ELEMENTS: and child.tagName in RECURSE_INTO_PARA_CONTAINERS:
#
fixup_paras_helper(doc, child) fixup_paras_helper(doc, child)
descriptions = child.getElementsByTagName("description") descriptions = child.getElementsByTagName("description")
for description in descriptions: for description in descriptions:
if DEBUG_PARA_FIXER:
sys.stderr.write("-- Fixing up <description> element...\n")
fixup_paras_helper(doc, description) fixup_paras_helper(doc, description)
def fixup_paras_helper(doc, container): def fixup_paras_helper(doc, container, depth=0):
# document is already normalized # document is already normalized
children = container.childNodes children = container.childNodes
start = 0 start = 0
start_fixed = 0 while len(children) > start:
i = len(children) start = skip_leading_nodes(children, start)
SKIP_ELEMENTS = PARA_LEVEL_ELEMENTS + PARA_LEVEL_PRECEEDERS if start >= len(children):
if DEBUG_PARA_FIXER: break
sys.stderr.write("fixup_paras_helper() called on <%s>; %d, %d\n" #
% (container.tagName, start, i)) # Either paragraph material or something to recurse into:
if i > start: #
# the first [start:i] children shoudl be rewritten as <para> elements if (children[start].nodeType == xml.dom.core.ELEMENT) \
# start by breaking text nodes that contain \n\n+ into multiple nodes and (children[start].tagName in RECURSE_INTO_PARA_CONTAINERS):
nstart, i = skip_leading_nodes(container.childNodes, start, i) fixup_paras_helper(doc, children[start])
if i > nstart: start = skip_leading_nodes(children, start + 1)
build_para(doc, container, nstart, i) continue
fixup_paras_helper(doc, container) #
# paragraph material:
#
build_para(doc, container, start, len(children))
if DEBUG_PARA_FIXER and depth == 10:
sys.exit(1)
start = start + 1
def build_para(doc, parent, start, i): def build_para(doc, parent, start, i):
children = parent.childNodes children = parent.childNodes
# collect all children until \n\n+ is found in a text node or a
# PARA_LEVEL_ELEMENT is found.
after = start + 1 after = start + 1
have_last = 0 have_last = 0
BREAK_ELEMENTS = PARA_LEVEL_ELEMENTS + FIXUP_PARA_ELEMENTS BREAK_ELEMENTS = PARA_LEVEL_ELEMENTS + FIXUP_PARA_ELEMENTS
# Collect all children until \n\n+ is found in a text node or a
# member of BREAK_ELEMENTS is found.
for j in range(start, i): for j in range(start, i):
after = j + 1 after = j + 1
child = children[j] child = children[j]
@ -521,6 +564,9 @@ def build_para(doc, parent, start, i):
break break
else: else:
have_last = 1 have_last = 1
if (start + 1) > after:
raise ConversionError(
"build_para() could not identify content to turn into a paragraph")
if children[after - 1].nodeType == xml.dom.core.TEXT: if children[after - 1].nodeType == xml.dom.core.TEXT:
# we may need to split off trailing white space: # we may need to split off trailing white space:
child = children[after - 1] child = children[after - 1]
@ -528,66 +574,60 @@ def build_para(doc, parent, start, i):
if string.rstrip(data) != data: if string.rstrip(data) != data:
have_last = 0 have_last = 0
child.splitText(len(string.rstrip(data))) child.splitText(len(string.rstrip(data)))
children = parent.childNodes
para = doc.createElement("para") para = doc.createElement("para")
prev = None prev = None
indexes = range(start, after) indexes = range(start, after)
indexes.reverse() indexes.reverse()
for j in indexes: for j in indexes:
node = children[j] node = parent.childNodes[j]
parent.removeChild(node) parent.removeChild(node)
para.insertBefore(node, prev) para.insertBefore(node, prev)
prev = node prev = node
if have_last: if have_last:
parent.appendChild(para) parent.appendChild(para)
return len(parent.childNodes)
else: else:
parent.insertBefore(para, parent.childNodes[start]) parent.insertBefore(para, parent.childNodes[start])
return start + 1
def skip_leading_nodes(children, start, i): def skip_leading_nodes(children, start):
i = min(i, len(children)) """Return index into children of a node at which paragraph building should
begin or a recursive call to fixup_paras_helper() should be made (for
subsections, etc.).
When the return value >= len(children), we've built all the paras we can
from this list of children.
"""
i = len(children)
while i > start: while i > start:
# skip over leading comments and whitespace: # skip over leading comments and whitespace:
try:
child = children[start] child = children[start]
except IndexError:
sys.stderr.write(
"skip_leading_nodes() failed at index %d\n" % start)
raise
nodeType = child.nodeType nodeType = child.nodeType
if nodeType == xml.dom.core.COMMENT: if nodeType == xml.dom.core.TEXT:
start = start + 1
elif nodeType == xml.dom.core.TEXT:
data = child.data data = child.data
shortened = string.lstrip(data) shortened = string.lstrip(data)
if shortened: if shortened:
if data != shortened: if data != shortened:
# break into two nodes: whitespace and non-whitespace # break into two nodes: whitespace and non-whitespace
child.splitText(len(data) - len(shortened)) child.splitText(len(data) - len(shortened))
return start + 1, i + 1 return start + 1
break return start
# all whitespace, just skip # all whitespace, just skip
start = start + 1
elif nodeType == xml.dom.core.ELEMENT: elif nodeType == xml.dom.core.ELEMENT:
if child.tagName in PARA_LEVEL_ELEMENTS + PARA_LEVEL_PRECEEDERS: tagName = child.tagName
if tagName in RECURSE_INTO_PARA_CONTAINERS:
return start
if tagName not in PARA_LEVEL_ELEMENTS + PARA_LEVEL_PRECEEDERS:
return start
start = start + 1 start = start + 1
else: return start
break
else:
break
return start, i
def fixup_rfc_references(doc): def fixup_rfc_references(doc):
rfc_nodes = [] for rfcnode in find_all_elements(doc, "rfc"):
for child in doc.childNodes: rfcnode.appendChild(doc.createTextNode(
if child.nodeType == xml.dom.core.ELEMENT: "RFC " + rfcnode.getAttribute("num")))
kids = child.getElementsByTagName("rfc")
for k in kids:
rfc_nodes.append(k)
for rfc_node in rfc_nodes:
rfc_node.appendChild(doc.createTextNode(
"RFC " + rfc_node.getAttribute("num")))
def fixup_signatures(doc): def fixup_signatures(doc):
@ -596,6 +636,7 @@ def fixup_signatures(doc):
args = child.getElementsByTagName("args") args = child.getElementsByTagName("args")
for arg in args: for arg in args:
fixup_args(doc, arg) fixup_args(doc, arg)
arg.normalize()
args = child.getElementsByTagName("constructor-args") args = child.getElementsByTagName("constructor-args")
for arg in args: for arg in args:
fixup_args(doc, arg) fixup_args(doc, arg)
@ -618,6 +659,22 @@ def fixup_args(doc, arglist):
return fixup_args(doc, arglist) return fixup_args(doc, arglist)
def fixup_sectionauthors(doc):
for sectauth in find_all_elements(doc, "sectionauthor"):
section = sectauth.parentNode
section.removeChild(sectauth)
sectauth._node.name = "author"
sectauth.appendChild(doc.createTextNode(
sectauth.getAttribute("name")))
sectauth.removeAttribute("name")
after = section.childNodes[2]
title = section.childNodes[1]
if title.nodeType == xml.dom.core.ELEMENT and title.tagName != "title":
after = section.childNodes[0]
section.insertBefore(doc.createTextNode("\n "), after)
section.insertBefore(sectauth, after)
_token_rx = re.compile(r"[a-zA-Z][a-zA-Z0-9.-]*$") _token_rx = re.compile(r"[a-zA-Z][a-zA-Z0-9.-]*$")
def write_esis(doc, ofp, knownempty): def write_esis(doc, ofp, knownempty):
@ -669,6 +726,7 @@ def convert(ifp, ofp):
cleanup_synopses(doc) cleanup_synopses(doc)
normalize(doc) normalize(doc)
fixup_paras(doc) fixup_paras(doc)
fixup_sectionauthors(doc)
remap_element_names(doc, { remap_element_names(doc, {
"tableii": ("table", {"cols": "2"}), "tableii": ("table", {"cols": "2"}),
"tableiii": ("table", {"cols": "3"}), "tableiii": ("table", {"cols": "3"}),