Files
clear-linux-documentation/source/_scripts/_python/manpages/manpages.py
T
Kevin Putnam a6204a18cb Adds manpages to reference section of Clear Linux docs (#1110)
* Build manpages with "make man".

Signed-off-by: Kevin Putnam <kevin.putnam@intel.com>

* Adds "make man" to github workflows.

Signed-off-by: Kevin Putnam <kevin.putnam@intel.com>

* Remove shell script dependency.

Signed-off-by: Kevin Putnam <kevin.putnam@intel.com>

* Change git URLs to https.

Signed-off-by: Kevin Putnam <kevin.putnam@intel.com>

* Moved git clone to Makefile.

Signed-off-by: Kevin Putnam <kevin.putnam@intel.com>

* Fixed typo.

Signed-off-by: Kevin Putnam <kevin.putnam@intel.com>

* Adding pandoc install.

Signed-off-by: Kevin Putnam <kevin.putnam@intel.com>

* Moved to Makefile.

Signed-off-by: Kevin Putnam <kevin.putnam@intel.com>

* First commit toward programmatic creation of manpages.rst

Signed-off-by: Kevin Putnam <kevin.putnam@intel.com>

* Multiple updates:

1. man-pages.rst now generated by script.
2. Updated Makefile and make.bat with "man" and "clean-man" recipes.
3. Fixed typo in conf.py.
4. Many improvements to manpages.py

Signed-off-by: Kevin Putnam <kevin.putnam@intel.com>

* Updated readme for man pages and added man-pages.rst

Signed-off-by: Kevin Putnam <kevin.putnam@intel.com>
2020-05-01 12:23:21 -07:00

196 lines
7.4 KiB
Python

#
# manpages.py
#
# maintainer: intelkevinputnam
#
# usage: python3 manpages.py
#
# dependencies: 1. clone of Clear Linux documentation https://github.com/clearlinux/clear-linux-documentation
# 2. Makefile or make.bat to create directories and clone repositories (``make man`` to generate. ``make clean-man`` to clean up.)
#
# This script does 4 things:
#
# 1. Find and move reStructuredText versions of man pages (looks for manName.sectionNumber.fileExtension: mixer.1.rst) to reference/manpages directory of Clear Linux docs.
# 2. Massage markdown and reStructuredText man pages into a normalized format.
# 3. Add cross reference links wherever man pages reference each other.
# 4. Create manpages.rst in reference directory (already included in toctree of reference/index.rst)
#
#
import sys
import re
import subprocess
from os import listdir
from os.path import join, isfile
from shutil import copyfile
paths = ["clr-man-pages","clr-power-tweaks/man","clrtrust/man","micro-config-drive/docs/","mixer-tools/docs","swupd-client/docs","tallow/man","telemetrics-client/docs/man"]
pathToRefSection = "../../../reference/"
manPageRegex = '.[0-9]+.rst'
mdManPageRegex = '.[0-9]+.md'
mdBoldItalicRegex = '\\*\\*\\`[a-z-._]+\\`\\*\\*'
manFiles = []
manGroups = {}
manNamePerms = {}
seeAlsoRefs = {}
gitHubGroup = "https://github.com/clearlinux/"
TOC = "\n.. contents::\n :local:\n"
manPageRest = """.. _man-pages:
Man pages
#########
These pages are generated from `Clear Linux\\* tool
repositories <https://github.com/clearlinux>`__. Updated |today|.
"""
manTOC = """.. toctree::
:maxdepth: 1
"""
def getPages(paths):
for path in paths:
manGroups[path] = []
files = listdir(path)
for file in files:
if file.endswith(".rst"):
regex_found = re.search(manPageRegex,file)
if regex_found:
filePath = join(path,file)
copyfile(filePath,file)
manFiles.append(file)
manGroups[path].append(file)
addManNamePermutations(file)
elif file.endswith(".md"):
regex_found = re.search(mdManPageRegex,file)
if regex_found:
filePath = join(path,file)
rstFileName = processMDFile(file,filePath)
manFiles.append(rstFileName)
manGroups[path].append(rstFileName)
addManNamePermutations(rstFileName)
def processMDFile(fileName,filePath):
manName = fileName.rstrip('.md')
manSection = manName.split('.')[-1]
manName = manName.rstrip('.' + manSection)
lines = []
headerInsert = ""
with open(filePath,'r') as m:
lines = m.readlines()
if "SYNOPSIS" in lines[0]:
header='='*len(manName)
lines.insert(0,"# " + manName + "\n\n:Manual section: "+manSection+"\n\n")
index = 0
first = True
for line in lines:
newLine = line.replace("**`","`") #Fix some markdown formatting weirdness that doesn't translate to reST
newLine = newLine.replace("`**","`")
if newLine.startswith("#") and not newLine.startswith("##"): #Fix if all headers are first level
if first:
first = False
else:
newLine = newLine.replace("#","##")
if "===" in newLine: #fix rst style header that is actually description in some docs
newLine = ""
headerInsert = manName + "\n" + '='*len(manName) + "\n\n"
lines[index] = newLine
index += 1
lines.insert(0,headerInsert)
with open(filePath,'w') as md:
md.writelines(lines)
rstFilePath = filePath.replace(".md",".rst")
rstFileName = fileName.replace(".md",".rst")
command = "pandoc " + filePath + " -o " + rstFilePath
subprocess.run(command, shell=True)
if isfile(rstFilePath):
copyfile(rstFilePath,rstFileName)
return rstFileName
def addManNamePermutations(fileName): #Based on all the syntactical permutations of references to man pages in the documents.
(manName,subsection) = getNameAndSubsection(fileName)
manNames = []
manNames.append("``" + manName + "``\\(" + subsection +")") # ``mixer.init``\(1)
manNames.append("``" + manName + "``\\ (" + subsection +")") # ``mixer.init``\ (1)
manNames.append("**``" + manName + "(" + subsection + ")``**") # **`mixer.init(1)`**
manNames.append("``" + manName + "(" + subsection + ")``") # ``mixer.init(1)``
manNames.append("**" + manName + "(" + subsection + ")**") # **mixer.init(1)**
manNames.append("`" + manName + "(" + subsection + ")`") # `mixer.init(1)`
manNamePerms[(manName,subsection)] = manNames
def getNameAndSubsection(manFileName):
manName = manFileName.rstrip('.rst')
manSection = manName.split('.')[-1]
manName = manName.rstrip('.'+manSection)
return (manName,manSection)
def linkToMan(manName,manSection):
return "`" + manName + "(" + manSection + ") <" + manName + "." + manSection + ".html>`__"
def buildManName(name,section):
return name + "(" + section + ")"
def updateManPages():
#makeSeeAlsoReplacements()
for file in manFiles:
manFile = ""
with open(file,'r') as f:
manFile = f.read()
#manFile = addTOC(manFile,file) # Not convinced adding TOC adds value.
#
# Add linked cross referencing for all manpages discovered.
# 1. the manNamePerms dictionary is created once all of the man page source files are discovered
# 2. Each document is checked for each permutation.
# 3. When a permuation is found it is replaced with a normalized version
# 4. Once normalized it is turned into a reStructuredText link (def linkToMan)
#
for nameAndSection, listOfPerms in manNamePerms.items():
for perm in listOfPerms:
manFile = manFile.replace(perm,buildManName(nameAndSection[0],nameAndSection[1]))
for manName, doNotUse in manNamePerms.items():
manFile = manFile.replace(buildManName(manName[0],manName[1]),linkToMan(manName[0],manName[1]))
with open(file,'w') as w:
w.write(manFile)
def addTOC(manContent,file):
# Find the first instance of the manpage Name
# Skip the underline of the header
# Add a new line
# Add contents directive with local modifier
# Add a new line
manSectionMeta = ":Manual section:"
manContentLines = manContent.split('\n')
index = 0
for line in manContentLines:
if manSectionMeta in line:
manContentLines.insert(index + 2,TOC)
break
index += 1
output = ""
for line in manContentLines:
output = output + line + "\n"
return output
def createManpagesRST():
filePath = join(pathToRefSection,"man-pages.rst")
with open(filePath,'w') as f:
manGrouping = manPageRest
for path, fileList in manGroups.items():
repoName = path.split("/")[0]
repoLink = gitHubGroup + repoName
repoReST = "`" + repoName + " <" + repoLink + ">`__"
manGrouping += repoReST + "\n"
manGrouping += "="*len(repoReST) + "\n\n"
manGrouping += manTOC
for file in fileList:
manGrouping += " manpages/" + file + "\n"
manGrouping += "\n"
f.write(manGrouping)
getPages(paths)
updateManPages()
createManpagesRST()