GUI: Regex builder for removing header and footer.

src/calibre/gui2/convert/__init__.py

@@ -77,6 +77,7 @@ class Widget(QWidget):
 
     def get_value(self, g):
         from calibre.gui2.convert.xpath_wizard import XPathEdit
+        from calibre.gui2.convert.regex_builder import RegexEdit
         ret = self.get_value_handler(g)
         if ret != 'this is a dummy return value, xcswx1avcx4x':
             return ret
@@ -94,12 +95,15 @@ class Widget(QWidget):
             return bool(g.isChecked())
         elif isinstance(g, XPathEdit):
             return g.xpath if g.xpath else None
+        elif isinstance(g, RegexEdit):
+            return g.regex if g.regex else None
         else:
             raise Exception('Can\'t get value from %s'%type(g))
 
 
     def set_value(self, g, val):
         from calibre.gui2.convert.xpath_wizard import XPathEdit
+        from calibre.gui2.convert.regex_builder import RegexEdit
         if self.set_value_handler(g, val):
             return
         if isinstance(g, (QSpinBox, QDoubleSpinBox)):
@@ -116,7 +120,7 @@ class Widget(QWidget):
             g.setCurrentIndex(idx)
         elif isinstance(g, QCheckBox):
             g.setCheckState(Qt.Checked if bool(val) else Qt.Unchecked)
-        elif isinstance(g, XPathEdit):
+        elif isinstance(g, (XPathEdit, RegexEdit)):
             g.edit.setText(val if val else '')
         else:
             raise Exception('Can\'t set value %s in %s'%(repr(val),

src/calibre/gui2/convert/regex_builder.py

@@ -0,0 +1,138 @@
+# -*- coding: utf-8 -*-
+
+__license__ = 'GPL 3'
+__copyright__ = '2009, John Schember <john@nachtimwald.com>'
+__docformat__ = 'restructuredtext en'
+
+import re
+
+from PyQt4.QtCore import SIGNAL, Qt
+from PyQt4.QtGui import QDialog, QWidget, QDialogButtonBox, QFileDialog, \
+    QBrush, QSyntaxHighlighter, QTextCharFormat
+
+from calibre.gui2.convert.regex_builder_ui import Ui_RegexBuilder
+from calibre.gui2.convert.xexp_edit_ui import Ui_Form as Ui_Edit
+from calibre.gui2 import qstring_to_unicode
+from calibre.gui2 import error_dialog
+from calibre.ebooks.oeb.iterator import EbookIterator
+from calibre.gui2.dialogs.choose_format import ChooseFormatDialog
+
+class RegexHighlighter(QSyntaxHighlighter):
+
+    def __init__(self, *args):
+        QSyntaxHighlighter.__init__(self, *args)
+
+        self.regex = u''
+
+    def update_regex(self, regex):
+        self.regex = qstring_to_unicode(regex)
+        self.rehighlight()
+
+    def highlightBlock(self, text):
+        valid_regex = True
+        text = qstring_to_unicode(text)
+        format = QTextCharFormat()
+        format.setBackground(QBrush(Qt.yellow))
+
+        if self.regex:
+            try:
+                for mo in re.finditer(self.regex, text):
+                    self.setFormat(mo.start(), mo.end() - mo.start(), format)
+            except:
+                valid_regex = False
+        self.emit(SIGNAL('regex_valid(PyQt_PyObject)'), valid_regex)
+
+class RegexBuilder(QDialog, Ui_RegexBuilder):
+
+    def __init__(self, db, book_id, regex, *args):
+        QDialog.__init__(self, *args)
+        self.setupUi(self)
+
+        self.regex.setText(regex)
+        self.regex_valid(True)
+        self.highlighter = RegexHighlighter(self.preview.document())
+
+        if not db or not book_id:
+            self.button_box.addButton(QDialogButtonBox.Open)
+        else:
+            self.select_format(db, book_id)
+
+        self.connect(self.button_box, SIGNAL('clicked(QAbstractButton*)'), self.button_clicked)
+        self.connect(self.regex, SIGNAL('textChanged(QString)'), self.highlighter.update_regex)
+        self.connect(self.highlighter, SIGNAL('regex_valid(PyQt_PyObject)'), self.regex_valid)
+
+    def regex_valid(self, valid):
+        if valid:
+            self.regex.setStyleSheet('QLineEdit { color: black; background-color: white; }')
+        else:
+            self.regex.setStyleSheet('QLineEdit { color: black; background-color: rgb(255,0,0,20%); }')
+
+    def select_format(self, db, book_id):
+        format = None
+        formats = db.formats(book_id, index_is_id=True).upper().split(',')
+        if len(formats) == 1:
+            format = formats[0]
+        elif len(formats) > 1:
+            d = ChooseFormatDialog(self, _('Choose the format to view'), formats)
+            d.exec_()
+            if d.result() == QDialog.Accepted:
+                format = d.format()
+
+        if not format:
+            error_dialog(self, _('No formats available'), _('Cannot build regex using the GUI builder without a book.'))
+            QDialog.reject()
+        else:
+            self.open_book(db.format_abspath(book_id, format, index_is_id=True))
+
+    def open_book(self, pathtoebook):
+        self.iterator = EbookIterator(pathtoebook)
+        self.iterator.__enter__()
+        text = [u'']
+        for path in self.iterator.spine:
+            html = open(path, 'rb').read().decode(path.encoding, 'replace')
+            text.append(html)
+        self.preview.setPlainText('\n\n'.join(text))
+
+    def button_clicked(self, button):
+        if button == self.button_box.button(QDialogButtonBox.Open):
+            name = QFileDialog.getOpenFileName(self, _('Open book'), _('~'))
+            if name:
+                self.open_book(qstring_to_unicode(name))
+        if button == self.button_box.button(QDialogButtonBox.Ok):
+            self.accept()
+
+class RegexEdit(QWidget, Ui_Edit):
+
+    def __init__(self, parent=None):
+        QWidget.__init__(self, parent)
+        self.setupUi(self)
+
+        self.book_id = None
+        self.db = None
+
+        self.connect(self.button, SIGNAL('clicked()'), self.builder)
+
+    def builder(self):
+        bld = RegexBuilder(self.db, self.book_id, self.edit.text(), self)
+        if bld.exec_() == bld.Accepted:
+            self.edit.setText(bld.regex.text())
+
+    def set_msg(self, msg):
+        self.msg.setText(msg)
+
+    def set_book_id(self, book_id):
+        self.book_id = book_id
+
+    def set_db(self, db):
+        self.db = db
+
+    @property
+    def text(self):
+        return unicode(self.edit.text())
+
+    @property
+    def regex(self):
+        return self.text
+
+    def check(self):
+        return True

src/calibre/gui2/convert/regex_builder.ui

@@ -0,0 +1,80 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<ui version="4.0">
+ <class>RegexBuilder</class>
+ <widget class="QDialog" name="RegexBuilder">
+  <property name="geometry">
+   <rect>
+    <x>0</x>
+    <y>0</y>
+    <width>662</width>
+    <height>505</height>
+   </rect>
+  </property>
+  <property name="windowTitle">
+   <string>Regex Builder</string>
+  </property>
+  <layout class="QGridLayout" name="gridLayout">
+   <item row="1" column="0" colspan="2">
+    <widget class="QGroupBox" name="groupBox">
+     <property name="title">
+      <string>Preview</string>
+     </property>
+     <layout class="QVBoxLayout" name="verticalLayout">
+      <item>
+       <widget class="QPlainTextEdit" name="preview">
+        <property name="undoRedoEnabled">
+         <bool>false</bool>
+        </property>
+        <property name="readOnly">
+         <bool>true</bool>
+        </property>
+        <property name="textInteractionFlags">
+         <set>Qt::TextSelectableByMouse</set>
+        </property>
+       </widget>
+      </item>
+     </layout>
+    </widget>
+   </item>
+   <item row="2" column="1">
+    <widget class="QDialogButtonBox" name="button_box">
+     <property name="orientation">
+      <enum>Qt::Horizontal</enum>
+     </property>
+     <property name="standardButtons">
+      <set>QDialogButtonBox::Cancel|QDialogButtonBox::Ok</set>
+     </property>
+    </widget>
+   </item>
+   <item row="0" column="0">
+    <widget class="QLabel" name="label">
+     <property name="text">
+      <string>Regex:</string>
+     </property>
+    </widget>
+   </item>
+   <item row="0" column="1">
+    <widget class="QLineEdit" name="regex"/>
+   </item>
+  </layout>
+ </widget>
+ <resources/>
+ <connections>
+  <connection>
+   <sender>button_box</sender>
+   <signal>rejected()</signal>
+   <receiver>RegexBuilder</receiver>
+   <slot>reject()</slot>
+   <hints>
+    <hint type="sourcelabel">
+     <x>316</x>
+     <y>260</y>
+    </hint>
+    <hint type="destinationlabel">
+     <x>286</x>
+     <y>274</y>
+    </hint>
+   </hints>
+  </connection>
+ </connections>
+</ui>

src/calibre/gui2/convert/structure_detection.py

@@ -34,11 +34,18 @@ class StructureDetectionWidget(Widget, Ui_Form):
         self.opt_chapter.set_msg(_('Detect chapters at (XPath expression):'))
         self.opt_page_breaks_before.set_msg(_('Insert page breaks before '
             '(XPath expression):'))
+        self.opt_header_regex.set_msg(_('Header regular expression:'))
+        self.opt_header_regex.set_book_id(book_id)
+        self.opt_header_regex.set_db(db)
+        self.opt_footer_regex.set_msg(_('Footer regular expression:'))
+        self.opt_footer_regex.set_book_id(book_id)
+        self.opt_footer_regex.set_db(db)
+
     def pre_commit_check(self):
         for x in ('header_regex', 'footer_regex'):
             x = getattr(self, 'opt_'+x)
             try:
-                pat = unicode(x.text())
+                pat = unicode(x.regex)
                 re.compile(pat)
             except Exception, err:
                 error_dialog(self, _('Invalid regular expression'),

src/calibre/gui2/convert/structure_detection.ui

@@ -28,8 +28,7 @@
     </widget>
    </item>
    <item row="1" column="1">
-    <widget class="QComboBox" name="opt_chapter_mark">
+    <widget class="QComboBox" name="opt_chapter_mark"/>
-    </widget>
    </item>
    <item row="2" column="0">
     <widget class="QCheckBox" name="opt_remove_first_image">
@@ -45,27 +44,17 @@
      </property>
     </widget>
    </item>
-   <item row="8" column="0">
+   <item row="8" column="0" colspan="2">
-    <widget class="QLabel" name="label_3">
-     <property name="text">
-      <string>&amp;Footer regular expression:</string>
-     </property>
-     <property name="buddy">
-      <cstring>opt_footer_regex</cstring>
-     </property>
-    </widget>
-   </item>
-   <item row="10" column="0" colspan="2">
     <widget class="QCheckBox" name="opt_preprocess_html">
      <property name="text">
       <string>&amp;Preprocess input file to possibly improve structure detection</string>
      </property>
     </widget>
    </item>
-   <item row="11" column="0" colspan="2">
+   <item row="9" column="0" colspan="2">
     <widget class="XPathEdit" name="opt_page_breaks_before" native="true"/>
    </item>
-   <item row="12" column="0">
+   <item row="10" column="0">
     <spacer name="verticalSpacer">
      <property name="orientation">
       <enum>Qt::Vertical</enum>
@@ -78,17 +67,7 @@
      </property>
     </spacer>
    </item>
-   <item row="5" column="0">
+   <item row="6" column="0">
-    <widget class="QLabel" name="label_2">
-     <property name="text">
-      <string>&amp;Header regular expression:</string>
-     </property>
-     <property name="buddy">
-      <cstring>opt_header_regex</cstring>
-     </property>
-    </widget>
-   </item>
-   <item row="7" column="0">
     <widget class="QCheckBox" name="opt_remove_footer">
      <property name="text">
       <string>Remove F&amp;ooter</string>
@@ -102,11 +81,11 @@
      </property>
     </widget>
    </item>
-   <item row="9" column="0" colspan="2">
+   <item row="5" column="0" colspan="2">
-    <widget class="QLineEdit" name="opt_footer_regex"/>
+    <widget class="RegexEdit" name="opt_header_regex" native="true"/>
    </item>
-   <item row="6" column="0" colspan="2">
+   <item row="7" column="0" colspan="2">
-    <widget class="QLineEdit" name="opt_header_regex"/>
+    <widget class="RegexEdit" name="opt_footer_regex" native="true"/>
    </item>
   </layout>
  </widget>
@@ -117,6 +96,12 @@
    <header>convert/xpath_wizard.h</header>
    <container>1</container>
   </customwidget>
+  <customwidget>
+   <class>RegexEdit</class>
+   <extends>QWidget</extends>
+   <header>regex_builder.h</header>
+   <container>1</container>
+  </customwidget>
  </customwidgets>
  <resources/>
  <connections/>

src/calibre/gui2/convert/xpath_wizard.py

@@ -9,7 +9,7 @@ __docformat__ = 'restructuredtext en'
 from PyQt4.Qt import QDialog, QWidget, SIGNAL, Qt, QDialogButtonBox, QVBoxLayout
 
 from calibre.gui2.convert.xpath_wizard_ui import Ui_Form
-from calibre.gui2.convert.xpath_edit_ui import Ui_Form as Ui_Edit
+from calibre.gui2.convert.xexp_edit_ui import Ui_Form as Ui_Edit
 
 
 class WizardWidget(QWidget, Ui_Form):