Sun, 16 May 2021 20:07:24 +0200
Replaced all imports of PyQt5 to PyQt6 and started to replace code using obsoleted methods and adapt to the PyQt6 enum usage.
# -*- coding: utf-8 -*- # Copyright (c) 2014 - 2021 Detlev Offenbach <detlev@die-offenbachs.de> # """ Module implementing a previewer widget for HTML, Markdown and ReST files. """ import os import threading import re import shutil import tempfile import sys import io import contextlib from PyQt6.QtCore import pyqtSlot, pyqtSignal, Qt, QUrl, QThread from PyQt6.QtGui import QCursor from PyQt6.QtWidgets import ( QWidget, QVBoxLayout, QLabel, QCheckBox, QSizePolicy, QToolTip ) from E5Gui.E5Application import e5App import Utilities import Preferences class PreviewerHTML(QWidget): """ Class implementing a previewer widget for HTML, Markdown and ReST files. """ def __init__(self, parent=None): """ Constructor @param parent reference to the parent widget (QWidget) """ super().__init__(parent) self.__layout = QVBoxLayout(self) self.titleLabel = QLabel(self) self.titleLabel.setWordWrap(True) self.titleLabel.setTextInteractionFlags( Qt.TextInteractionFlag.NoTextInteraction) self.__layout.addWidget(self.titleLabel) self.__previewAvailable = True try: from PyQt6.QtWebEngineWidgets import QWebEngineView self.previewView = QWebEngineView(self) self.previewView.page().linkHovered.connect(self.__showLink) except ImportError: self.__previewAvailable = False self.titleLabel.setText(self.tr( "<b>HTML Preview is not available!<br/>" "Install PyQtWebEngine.</b>")) self.titleLabel.setAlignment(Qt.AlignmentFlag.AlignHCenter) self.__layout.addStretch() return sizePolicy = QSizePolicy(QSizePolicy.Policy.Preferred, QSizePolicy.Policy.Expanding) sizePolicy.setHorizontalStretch(0) sizePolicy.setVerticalStretch(0) sizePolicy.setHeightForWidth( self.previewView.sizePolicy().hasHeightForWidth()) self.previewView.setSizePolicy(sizePolicy) self.previewView.setContextMenuPolicy( Qt.ContextMenuPolicy.NoContextMenu) self.previewView.setUrl(QUrl("about:blank")) self.__layout.addWidget(self.previewView) self.jsCheckBox = QCheckBox(self.tr("Enable JavaScript"), self) self.jsCheckBox.setToolTip(self.tr( "Select to enable JavaScript for HTML previews")) self.__layout.addWidget(self.jsCheckBox) self.ssiCheckBox = QCheckBox(self.tr("Enable Server Side Includes"), self) self.ssiCheckBox.setToolTip(self.tr( "Select to enable support for Server Side Includes")) self.__layout.addWidget(self.ssiCheckBox) self.jsCheckBox.clicked[bool].connect(self.on_jsCheckBox_clicked) self.ssiCheckBox.clicked[bool].connect(self.on_ssiCheckBox_clicked) self.previewView.titleChanged.connect(self.on_previewView_titleChanged) self.jsCheckBox.setChecked( Preferences.getUI("ShowFilePreviewJS")) self.ssiCheckBox.setChecked( Preferences.getUI("ShowFilePreviewSSI")) self.__scrollBarPositions = {} self.__vScrollBarAtEnd = {} self.__hScrollBarAtEnd = {} self.__processingThread = PreviewProcessingThread() self.__processingThread.htmlReady.connect(self.__setHtml) self.__previewedPath = None self.__previewedEditor = None def shutdown(self): """ Public method to perform shutdown actions. """ if self.__previewAvailable: self.__processingThread.wait() @pyqtSlot(bool) def on_jsCheckBox_clicked(self, checked): """ Private slot to enable/disable JavaScript. @param checked state of the checkbox (boolean) """ Preferences.setUI("ShowFilePreviewJS", checked) self.__setJavaScriptEnabled(checked) def __setJavaScriptEnabled(self, enable): """ Private method to enable/disable JavaScript. @param enable flag indicating the enable state (boolean) """ self.jsCheckBox.setChecked(enable) settings = self.previewView.settings() settings.setAttribute(settings.JavascriptEnabled, enable) self.processEditor() @pyqtSlot(bool) def on_ssiCheckBox_clicked(self, checked): """ Private slot to enable/disable SSI. @param checked state of the checkbox (boolean) """ Preferences.setUI("ShowFilePreviewSSI", checked) self.processEditor() @pyqtSlot(str) def __showLink(self, urlStr): """ Private slot to show the hovered link in a tooltip. @param urlStr hovered URL @type str """ QToolTip.showText(QCursor.pos(), urlStr, self.previewView) def processEditor(self, editor=None): """ Public slot to process an editor's text. @param editor editor to be processed (Editor) """ if not self.__previewAvailable: return if editor is None: editor = self.__previewedEditor else: self.__previewedEditor = editor if editor is not None: fn = editor.getFileName() if fn: extension = os.path.normcase(os.path.splitext(fn)[1][1:]) else: extension = "" if ( extension in Preferences.getEditor( "PreviewHtmlFileNameExtensions") or editor.getLanguage() == "HTML" ): language = "HTML" elif ( extension in Preferences.getEditor( "PreviewMarkdownFileNameExtensions") or editor.getLanguage().lower() == "markdown" ): language = "Markdown" elif ( extension in Preferences.getEditor( "PreviewRestFileNameExtensions") or editor.getLanguage().lower() == "restructuredtext" ): language = "ReST" else: self.__setHtml(fn, self.tr( "<p>No preview available for this type of file.</p>")) return if fn: project = e5App().getObject("Project") if project.isProjectFile(fn): rootPath = project.getProjectPath() else: rootPath = os.path.dirname(os.path.abspath(fn)) else: rootPath = "" if bool(editor.text()): self.__processingThread.process( fn, language, editor.text(), self.ssiCheckBox.isChecked(), rootPath, Preferences.getEditor("PreviewRestUseSphinx"), Preferences.getEditor("PreviewMarkdownNLtoBR"), Preferences.getEditor( "PreviewMarkdownUsePyMdownExtensions"), Preferences.getEditor("PreviewMarkdownHTMLFormat"), Preferences.getEditor("PreviewRestDocutilsHTMLFormat")) def __setHtml(self, filePath, html, rootPath): """ Private method to set the HTML to the view and restore the scroll bars positions. @param filePath file path of the previewed editor @type str @param html processed HTML text ready to be shown @type str @param rootPath path of the web site root @type str """ self.__previewedPath = Utilities.normcasepath( Utilities.fromNativeSeparators(filePath)) self.__saveScrollBarPositions() self.previewView.page().loadFinished.connect( self.__restoreScrollBarPositions) if not filePath: filePath = "/" baseUrl = ( QUrl.fromLocalFile(rootPath + "/index.html") if rootPath else QUrl.fromLocalFile(filePath) ) self.previewView.setHtml(html, baseUrl=baseUrl) if self.__previewedEditor: self.__previewedEditor.setFocus() @pyqtSlot(str) def on_previewView_titleChanged(self, title): """ Private slot to handle a change of the title. @param title new title (string) """ if title: self.titleLabel.setText(self.tr("Preview - {0}").format(title)) else: self.titleLabel.setText(self.tr("Preview")) def __saveScrollBarPositions(self): """ Private method to save scroll bar positions for a previewed editor. """ from PyQt6.QtCore import QPoint try: pos = self.previewView.scrollPosition() except AttributeError: pos = self.__execJavaScript( "(function() {" "var res = {" " x: 0," " y: 0," "};" "res.x = window.scrollX;" "res.y = window.scrollY;" "return res;" "})()" ) pos = QPoint(0, 0) if pos is None else QPoint(pos["x"], pos["y"]) self.__scrollBarPositions[self.__previewedPath] = pos self.__hScrollBarAtEnd[self.__previewedPath] = False self.__vScrollBarAtEnd[self.__previewedPath] = False def __restoreScrollBarPositions(self): """ Private method to restore scroll bar positions for a previewed editor. """ if self.__previewedPath not in self.__scrollBarPositions: return pos = self.__scrollBarPositions[self.__previewedPath] self.previewView.page().runJavaScript( "window.scrollTo({0}, {1});".format(pos.x(), pos.y())) def __execJavaScript(self, script): """ Private function to execute a JavaScript function Synchroneously. @param script JavaScript script source to be executed @type str @return result of the script @rtype depending upon script result """ from PyQt6.QtCore import QEventLoop loop = QEventLoop() resultDict = {"res": None} def resultCallback(res, resDict=resultDict): if loop and loop.isRunning(): resDict["res"] = res loop.quit() self.previewView.page().runJavaScript( script, resultCallback) loop.exec() return resultDict["res"] class PreviewProcessingThread(QThread): """ Class implementing a thread to process some text into HTML usable by the previewer view. @signal htmlReady(str, str, str) emitted with the file name, the processed HTML and the web site root path to signal the availability of the processed HTML """ htmlReady = pyqtSignal(str, str, str) def __init__(self, parent=None): """ Constructor @param parent reference to the parent object (QObject) """ super().__init__() self.__lock = threading.Lock() def process(self, filePath, language, text, ssiEnabled, rootPath, useSphinx, convertNewLineToBreak, usePyMdownExtensions, markdownHtmlFormat, restDocutilsHtmlFormat): """ Public method to convert the given text to HTML. @param filePath file path of the text @type str @param language language of the text @type str @param text text to be processed @type str @param ssiEnabled flag indicating to do some (limited) SSI processing @type bool @param rootPath root path to be used for SSI processing @type str @param useSphinx flag indicating to use Sphinx to generate the ReST preview @type bool @param convertNewLineToBreak flag indicating to convert new lines to HTML break (Markdown only) @type bool @param usePyMdownExtensions flag indicating to enable the PyMdown extensions, if they are available @type bool @param markdownHtmlFormat HTML format to be generated by markdown @type str @param restDocutilsHtmlFormat HTML format to be generated by docutils @type str """ with self.__lock: self.__filePath = filePath self.__language = language self.__text = text self.__ssiEnabled = ssiEnabled self.__rootPath = rootPath self.__haveData = True self.__useSphinx = useSphinx self.__convertNewLineToBreak = convertNewLineToBreak self.__usePyMdownExtensions = usePyMdownExtensions self.__markdownHtmlFormat = markdownHtmlFormat self.__restDocutilsHtmlFormat = restDocutilsHtmlFormat if not self.isRunning(): self.start(QThread.Priority.LowPriority) def run(self): """ Public thread method to convert the stored data. """ while True: # exits with break with self.__lock: filePath = self.__filePath language = self.__language text = self.__text ssiEnabled = self.__ssiEnabled rootPath = self.__rootPath useSphinx = self.__useSphinx convertNewLineToBreak = self.__convertNewLineToBreak usePyMdownExtensions = self.__usePyMdownExtensions markdownHtmlFormat = self.__markdownHtmlFormat restDocutilsHtmlFormat = self.__restDocutilsHtmlFormat self.__haveData = False html = self.__getHtml(language, text, ssiEnabled, filePath, rootPath, useSphinx, convertNewLineToBreak, usePyMdownExtensions, markdownHtmlFormat, restDocutilsHtmlFormat) with self.__lock: if not self.__haveData: self.htmlReady.emit(filePath, html, rootPath) break # else - next iteration def __getHtml(self, language, text, ssiEnabled, filePath, rootPath, useSphinx, convertNewLineToBreak, usePyMdownExtensions, markdownHtmlFormat, restDocutilsHtmlFormat): """ Private method to process the given text depending upon the given language. @param language language of the text @type str @param text to be processed @type str @param ssiEnabled flag indicating to do some (limited) SSI processing @type bool @param filePath file path of the text @type str @param rootPath root path to be used for SSI processing @type str @param useSphinx flag indicating to use Sphinx to generate the ReST preview @type bool @param convertNewLineToBreak flag indicating to convert new lines to HTML break (Markdown only) @type bool @param usePyMdownExtensions flag indicating to enable the PyMdown extensions, if they are available @type bool @param markdownHtmlFormat HTML format to be generated by markdown @type str @param restDocutilsHtmlFormat HTML format to be generated by docutils @type str @return processed HTML text @rtype str """ if language == "HTML": if ssiEnabled: html = self.__processSSI(text, filePath, rootPath) else: html = text return self.__processRootPath(html, rootPath) elif language == "Markdown": return self.__convertMarkdown( text, convertNewLineToBreak, usePyMdownExtensions, markdownHtmlFormat) elif language == "ReST": return self.__convertReST(text, useSphinx, restDocutilsHtmlFormat) else: return self.tr( "<p>No preview available for this type of file.</p>") def __processSSI(self, txt, filename, root): """ Private method to process the given text for SSI statements. Note: Only a limited subset of SSI statements are supported. @param txt text to be processed (string) @param filename name of the file associated with the given text (string) @param root directory of the document root (string) @return processed HTML (string) """ if not filename: return txt # SSI include incRe = re.compile( r"""<!--#include[ \t]+(virtual|file)=[\"']([^\"']+)[\"']\s*-->""", re.IGNORECASE) baseDir = os.path.dirname(os.path.abspath(filename)) docRoot = root if root != "" else baseDir while True: incMatch = incRe.search(txt) if incMatch is None: break if incMatch.group(1) == "virtual": incFile = Utilities.normjoinpath(docRoot, incMatch.group(2)) elif incMatch.group(1) == "file": incFile = Utilities.normjoinpath(baseDir, incMatch.group(2)) else: incFile = "" if os.path.exists(incFile): try: with open(incFile, "r") as f: incTxt = f.read() except OSError: # remove SSI include incTxt = "" else: # remove SSI include incTxt = "" txt = txt[:incMatch.start(0)] + incTxt + txt[incMatch.end(0):] return txt def __processRootPath(self, txt, root): """ Private method to adjust absolute references to the given root path. @param txt text to be processed @type str @param root directory of the document root @type str @return processed HTML @rtype str """ if not root: return txt root = Utilities.fromNativeSeparators(root) if not root.endswith("/"): root += "/" rootLen = len(root) refRe = re.compile( r"""(href|src)=[\\"']/([^\\"']+)[\\"']""", re.IGNORECASE) pos = 0 while True: refMatch = refRe.search(txt, pos) if refMatch is None: break txt = (txt[:refMatch.start(0)] + refMatch.group(1) + '="' + root + refMatch.group(2) + '"' + txt[refMatch.end(0):]) pos = refMatch.end(0) + rootLen return txt def __convertReST(self, text, useSphinx, restDocutilsHtmlFormat): """ Private method to convert ReST text into HTML. @param text text to be processed (string) @param useSphinx flag indicating to use Sphinx to generate the ReST preview (boolean) @param restDocutilsHtmlFormat HTML format to be generated by docutils (string) @return processed HTML (string) """ if useSphinx: return self.__convertReSTSphinx(text) else: return self.__convertReSTDocutils(text, restDocutilsHtmlFormat) def __convertReSTSphinx(self, text): """ Private method to convert ReST text into HTML using 'sphinx'. @param text text to be processed (string) @return processed HTML (string) """ try: from sphinx.application import Sphinx # __IGNORE_EXCEPTION__ except ImportError: return self.tr( """<p>ReStructuredText preview requires the""" """ <b>sphinx</b> package.<br/>Install it with""" """ your package manager,'pip install Sphinx' or see""" """ <a href="http://pypi.python.org/pypi/Sphinx">""" """this page.</a></p>""" """<p>Alternatively you may disable Sphinx usage""" """ on the Editor, Filehandling configuration page.</p>""") srcTempDir = tempfile.mkdtemp(prefix="eric-rest-src-") outTempDir = tempfile.mkdtemp(prefix="eric-rest-out-") doctreeTempDir = tempfile.mkdtemp(prefix="eric-rest-doctree-") try: filename = 'sphinx_preview' basePath = os.path.join(srcTempDir, filename) with open(basePath + '.rst', 'w', encoding='utf-8') as fh: fh.write(text) overrides = {'html_add_permalinks': False, 'html_copy_source': False, 'html_title': 'Sphinx preview', 'html_use_index': False, 'html_use_modindex': False, 'html_use_smartypants': True, 'master_doc': filename} app = Sphinx(srcdir=srcTempDir, confdir=None, outdir=outTempDir, doctreedir=doctreeTempDir, buildername='html', confoverrides=overrides, status=None, warning=io.StringIO()) app.build(force_all=True, filenames=None) basePath = os.path.join(outTempDir, filename) with open(basePath + '.html', 'r', encoding='utf-8') as fh: html = fh.read() finally: shutil.rmtree(srcTempDir) shutil.rmtree(outTempDir) shutil.rmtree(doctreeTempDir) return html def __convertReSTDocutils(self, text, htmlFormat): """ Private method to convert ReST text into HTML using 'docutils'. @param text text to be processed (string) @param htmlFormat HTML format to be generated (string) @return processed HTML (string) """ if 'sphinx' in sys.modules: # Make sure any Sphinx polution of docutils has been removed. unloadKeys = [k for k in sys.modules.keys() if k.startswith(('docutils', 'sphinx'))] for key in unloadKeys: sys.modules.pop(key) try: import docutils.core # __IGNORE_EXCEPTION__ import docutils.utils # __IGNORE_EXCEPTION__ except ImportError: return self.tr( """<p>ReStructuredText preview requires the""" """ <b>python-docutils</b> package.<br/>Install it with""" """ your package manager, 'pip install docutils' or see""" """ <a href="http://pypi.python.org/pypi/docutils">""" """this page.</a></p>""") # redirect sys.stderr because we are not interested in it here origStderr = sys.stderr sys.stderr = io.StringIO() try: html = docutils.core.publish_string( text, writer_name=htmlFormat.lower()).decode("utf-8") except docutils.utils.SystemMessage as err: errStr = str(err).split(":")[-1].replace("\n", "<br/>") return self.tr( """<p>Docutils returned an error:</p><p>{0}</p>""" ).format(errStr) sys.stderr = origStderr return html def __convertMarkdown(self, text, convertNewLineToBreak, usePyMdownExtensions, htmlFormat): """ Private method to convert Markdown text into HTML. @param text text to be processed @type str @param convertNewLineToBreak flag indicating to convert new lines to HTML break (Markdown only) @type bool @param usePyMdownExtensions flag indicating to enable the PyMdown extensions, if they are available @type bool @param htmlFormat HTML format to be generated by markdown @type str @return processed HTML @rtype str """ try: import markdown # __IGNORE_EXCEPTION__ except ImportError: return self.tr( """<p>Markdown preview requires the <b>Markdown</b> """ """package.<br/>Install it with your package manager,""" """ 'pip install Markdown' or see """ """<a href="http://pythonhosted.org/Markdown/install.html">""" """installation instructions.</a></p>""") from . import PreviewerHTMLStyles from . import MarkdownExtensions extensions = [] mermaidNeeded = False if ( Preferences.getEditor("PreviewMarkdownMermaid") and MarkdownExtensions.MermaidRegexFullText.search(text) ): extensions.append(MarkdownExtensions.MermaidExtension()) mermaidNeeded = True if convertNewLineToBreak: extensions.append('nl2br') pyMdown = False if usePyMdownExtensions: with contextlib.suppress(ImportError): import pymdownx # __IGNORE_EXCEPTION__ __IGNORE_WARNING__ # PyPI package is 'pymdown-extensions' extensions.extend([ 'toc', 'pymdownx.extra', 'pymdownx.caret', 'pymdownx.emoji', 'pymdownx.mark', 'pymdownx.tilde', 'pymdownx.keys', 'pymdownx.tasklist', 'pymdownx.smartsymbols', ]) pyMdown = True if not pyMdown: extensions.extend(['extra', 'toc']) # version 2.0 supports only extension names, not instances if ( markdown.version_info[0] > 2 or (markdown.version_info[0] == 2 and markdown.version_info[1] > 0) ): extensions.append(MarkdownExtensions.SimplePatternExtension()) if Preferences.getEditor("PreviewMarkdownMathJax"): mathjax = ( "<script type='text/javascript' id='MathJax-script' async" " src='https://cdn.jsdelivr.net/npm/mathjax@3/es5/" "tex-chtml.js'>\n" "</script>\n" ) # prepare text for mathjax text = ( text .replace(r"\(", r"\\(") .replace(r"\)", r"\\)") .replace(r"\[", r"\\[") .replace(r"\]", r"\\]") ) else: mathjax = "" if mermaidNeeded: mermaid = ( "<script type='text/javascript' id='Mermaid-script'" " src='https://unpkg.com/mermaid@8/dist/mermaid.min.js'>\n" "</script>\n" ) if e5App().usesDarkPalette(): mermaid_initialize = ( "<script>mermaid.initialize({" "theme: 'dark', " "startOnLoad:true" "});</script>" ) else: mermaid_initialize = ( "<script>mermaid.initialize({" "theme: 'default', " "startOnLoad:true" "});</script>" ) else: mermaid = "" mermaid_initialize = "" htmlFormat = Preferences.getEditor("PreviewMarkdownHTMLFormat").lower() body = markdown.markdown(text, extensions=extensions, output_format=htmlFormat.lower()) style = ( (PreviewerHTMLStyles.css_markdown_dark + PreviewerHTMLStyles.css_pygments_dark) if e5App().usesDarkPalette() else (PreviewerHTMLStyles.css_markdown_light + PreviewerHTMLStyles.css_pygments_light) ) if htmlFormat == "xhtml1": head = ( '''<!DOCTYPE html PUBLIC "-//W3C//DTD''' ''' XHTML 1.0 Transitional//EN"\n''' ''' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional''' '''.dtd">\n''' '''<html xmlns="http://www.w3.org/1999/xhtml">\n''' ) elif htmlFormat == "html5": head = ( '''<!DOCTYPE html>\n''' '''<html lang="EN">\n''' ) else: head = '<html lang="EN">\n' head += '''<head>\n''' head += ( '''<meta name="Generator" content="eric" />\n''' '''<meta http-equiv="Content-Type" ''' '''content="text/html; charset=utf-8" />\n''' '''{0}''' '''{1}''' '''<style type="text/css">''' '''{2}''' '''</style>\n''' '''</head>\n''' '''<body>\n''' ).format(mathjax, mermaid, style) foot = '''\n</body>\n</html>\n''' return head + body + mermaid_initialize + foot