From: Vincent Vanwaelscappel <vincent@cubedesigners.com>
Date: Mon, 22 Aug 2022 20:01:11 +0000 (+0200)
Subject: wip #5410 @0.5
X-Git-Url: http://git.cubedesigners.com/?a=commitdiff_plain;h=dda06b6bcefe9ad329a0349bc1bfab4e0fbc910c;p=cubist_pdf.git

wip #5410 @0.5
---

diff --git a/.idea/misc.xml b/.idea/misc.xml
index b44ac24..ad68c41 100644
--- a/.idea/misc.xml
+++ b/.idea/misc.xml
@@ -3,7 +3,7 @@
   <component name="AhkProjectSettings">
     <option name="defaultAhkSdk" value="AutoHotkey" />
   </component>
-  <component name="ProjectRootManager" version="2" languageLevel="JDK_1_6" project-jdk-name="17" project-jdk-type="JavaSDK">
+  <component name="ProjectRootManager" version="2" languageLevel="JDK_1_6" project-jdk-name="18" project-jdk-type="JavaSDK">
     <output url="file://$PROJECT_DIR$/out" />
   </component>
 </project>
\ No newline at end of file
diff --git a/resources/tools/fwstk/.idea/misc.xml b/resources/tools/fwstk/.idea/misc.xml
index bec7cd5..e5dddc0 100644
--- a/resources/tools/fwstk/.idea/misc.xml
+++ b/resources/tools/fwstk/.idea/misc.xml
@@ -178,7 +178,7 @@
       </profile-state>
     </entry>
   </component>
-  <component name="ProjectRootManager" version="2" languageLevel="JDK_1_8" default="true" project-jdk-name="1.8" project-jdk-type="JavaSDK">
+  <component name="ProjectRootManager" version="2" languageLevel="JDK_18" default="true" project-jdk-name="18" project-jdk-type="JavaSDK">
     <output url="file://$PROJECT_DIR$/out" />
   </component>
   <component name="SvnBranchConfigurationManager">
diff --git a/resources/tools/fwstk/.idea/vcs.xml b/resources/tools/fwstk/.idea/vcs.xml
index 45cf85c..bc8238e 100644
--- a/resources/tools/fwstk/.idea/vcs.xml
+++ b/resources/tools/fwstk/.idea/vcs.xml
@@ -2,5 +2,6 @@
 <project version="4">
   <component name="VcsDirectoryMappings">
     <mapping directory="" vcs="svn" />
+    <mapping directory="$PROJECT_DIR$/../../.." vcs="Git" />
   </component>
 </project>
\ No newline at end of file
diff --git a/resources/tools/fwstk/.idea/workspace.xml b/resources/tools/fwstk/.idea/workspace.xml
index 082ba1c..3b4574a 100644
--- a/resources/tools/fwstk/.idea/workspace.xml
+++ b/resources/tools/fwstk/.idea/workspace.xml
@@ -7,10 +7,13 @@
   </component>
   <component name="ChangeListManager">
     <list default="true" id="f146bc67-2578-4de3-9db2-94d2d43e9e83" name="Default" comment="wip #4097 @0.25">
-      <change beforePath="$PROJECT_DIR$/.idea/runConfigurations.xml" beforeDir="false" />
+      <change beforePath="$PROJECT_DIR$/../../../.idea/misc.xml" beforeDir="false" afterPath="$PROJECT_DIR$/../../../.idea/misc.xml" afterDir="false" />
+      <change beforePath="$PROJECT_DIR$/.idea/misc.xml" beforeDir="false" afterPath="$PROJECT_DIR$/.idea/misc.xml" afterDir="false" />
+      <change beforePath="$PROJECT_DIR$/.idea/vcs.xml" beforeDir="false" afterPath="$PROJECT_DIR$/.idea/vcs.xml" afterDir="false" />
       <change beforePath="$PROJECT_DIR$/.idea/workspace.xml" beforeDir="false" afterPath="$PROJECT_DIR$/.idea/workspace.xml" afterDir="false" />
-      <change beforePath="$PROJECT_DIR$/out/artifacts/fwstk_jar/fwstk.jar" beforeDir="false" afterPath="$PROJECT_DIR$/out/artifacts/fwstk_jar/fwstk.jar" afterDir="false" />
-      <change beforePath="$PROJECT_DIR$/src/com/fluidbook/fwstk/Main.java" beforeDir="false" afterPath="$PROJECT_DIR$/src/com/fluidbook/fwstk/Main.java" afterDir="false" />
+      <change beforePath="$PROJECT_DIR$/src/com/fluidbook/fwstk/TextsThread.java" beforeDir="false" afterPath="$PROJECT_DIR$/src/com/fluidbook/fwstk/TextsThread.java" afterDir="false" />
+      <change beforePath="$PROJECT_DIR$/src/com/fluidbook/fwstk/layout/Group.java" beforeDir="false" afterPath="$PROJECT_DIR$/src/com/fluidbook/fwstk/layout/Group.java" afterDir="false" />
+      <change beforePath="$PROJECT_DIR$/../../../src/PDFTools.php" beforeDir="false" afterPath="$PROJECT_DIR$/../../../src/PDFTools.php" afterDir="false" />
     </list>
     <option name="SHOW_DIALOG" value="false" />
     <option name="HIGHLIGHT_CONFLICTS" value="true" />
@@ -66,17 +69,20 @@
       <dir>H:\Works\cubeExtranet\fluidbook\tools\fwstk\src</dir>
     </dirStrings>
   </component>
+  <component name="Git.Settings">
+    <option name="RECENT_GIT_ROOT_PATH" value="$PROJECT_DIR$/../../.." />
+  </component>
   <component name="HighlightingSettingsPerFile">
-    <setting file="file://$PROJECT_DIR$/src/com/fluidbook/fwstk/layout/LayoutStripper.java" root0="FORCE_HIGHLIGHTING" />
-    <setting file="file://$PROJECT_DIR$/src/com/fluidbook/fwstk/layout/Separator.java" root0="FORCE_HIGHLIGHTING" />
     <setting file="file://$PROJECT_DIR$/.idea/artifacts/fwstk_jar.xml" root0="FORCE_HIGHLIGHTING" />
-    <setting file="file://$PROJECT_DIR$/src/com/fluidbook/fwstk/Link.java" root0="FORCE_HIGHLIGHTING" />
     <setting file="file://$PROJECT_DIR$/.idea/copyright/profiles_settings.xml" root0="FORCE_HIGHLIGHTING" />
-    <setting file="file://$PROJECT_DIR$/src/com/fluidbook/fwstk/layout/Space.java" root0="FORCE_HIGHLIGHTING" />
     <setting file="file://$PROJECT_DIR$/MANIFEST.MF" root0="FORCE_HIGHLIGHTING" />
-    <setting file="file://$PROJECT_DIR$/src/com/fluidbook/fwstk/layout/Page.java" root0="FORCE_HIGHLIGHTING" />
     <setting file="file://$PROJECT_DIR$/fwstk.iml" root0="FORCE_HIGHLIGHTING" />
+    <setting file="file://$PROJECT_DIR$/src/com/fluidbook/fwstk/Link.java" root0="FORCE_HIGHLIGHTING" />
     <setting file="file://$PROJECT_DIR$/src/com/fluidbook/fwstk/Main.java" root0="FORCE_HIGHLIGHTING" />
+    <setting file="file://$PROJECT_DIR$/src/com/fluidbook/fwstk/layout/LayoutStripper.java" root0="FORCE_HIGHLIGHTING" />
+    <setting file="file://$PROJECT_DIR$/src/com/fluidbook/fwstk/layout/Page.java" root0="FORCE_HIGHLIGHTING" />
+    <setting file="file://$PROJECT_DIR$/src/com/fluidbook/fwstk/layout/Separator.java" root0="FORCE_HIGHLIGHTING" />
+    <setting file="file://$PROJECT_DIR$/src/com/fluidbook/fwstk/layout/Space.java" root0="FORCE_HIGHLIGHTING" />
   </component>
   <component name="MarkdownSettingsMigration">
     <option name="stateVersion" value="1" />
@@ -85,7 +91,7 @@
     <servers />
   </component>
   <component name="ProjectId" id="1S3xbFxrVuUoNhkuaZ10sq6V8Ta" />
-  <component name="ProjectLevelVcsManager">
+  <component name="ProjectLevelVcsManager" settingsEditedManually="true">
     <ConfirmationsSetting value="1" id="Add" />
     <ConfirmationsSetting value="2" id="Remove" />
   </component>
@@ -93,26 +99,22 @@
     <option name="hideEmptyMiddlePackages" value="true" />
     <option name="showLibraryContents" value="true" />
   </component>
-  <component name="PropertiesComponent">
-    <property name="AnalyzeApkAction.lastApkPath" value="$PROJECT_DIR$/out/artifacts/fwstk_jar" />
-    <property name="WebServerToolWindowFactoryState" value="true" />
-    <property name="WebServerToolWindowPanel.toolwindow.highlight.mappings" value="true" />
-    <property name="WebServerToolWindowPanel.toolwindow.highlight.symlinks" value="true" />
-    <property name="WebServerToolWindowPanel.toolwindow.show.date" value="false" />
-    <property name="WebServerToolWindowPanel.toolwindow.show.permissions" value="false" />
-    <property name="WebServerToolWindowPanel.toolwindow.show.size" value="false" />
-    <property name="aspect.path.notification.shown" value="true" />
-    <property name="com.android.tools.idea.instantapp.provision.ProvisionBeforeRunTaskProvider.myTimeStamp" value="1542740619620" />
-    <property name="js-jscs-nodeInterpreter" value="C:\Program Files\nodejs\node.exe" />
-    <property name="js.eslint.eslintPackage" value="" />
-    <property name="last_opened_file_path" value="$PROJECT_DIR$" />
-    <property name="nodejs_interpreter_path" value="C:/Program Files/nodejs/node" />
-    <property name="project.structure.last.edited" value="Modules" />
-    <property name="project.structure.proportion" value="0.15193799" />
-    <property name="project.structure.side.proportion" value="0.2" />
-    <property name="settings.editor.selected.configurable" value="configurable.group.language" />
-    <property name="vue.rearranger.settings.migration" value="true" />
-  </component>
+  <component name="PropertiesComponent"><![CDATA[{
+  "keyToString": {
+    "WebServerToolWindowFactoryState": "true",
+    "WebServerToolWindowPanel.toolwindow.highlight.mappings": "true",
+    "WebServerToolWindowPanel.toolwindow.highlight.symlinks": "true",
+    "WebServerToolWindowPanel.toolwindow.show.date": "false",
+    "WebServerToolWindowPanel.toolwindow.show.permissions": "false",
+    "WebServerToolWindowPanel.toolwindow.show.size": "false",
+    "last_opened_file_path": "D:/Works/cubist_pdf/resources/tools/fwstk",
+    "project.structure.last.edited": "Project",
+    "project.structure.proportion": "0.0",
+    "project.structure.side.proportion": "0.0",
+    "ruby.rails.projectView.checked": "true",
+    "vue.rearranger.settings.migration": "true"
+  }
+}]]></component>
   <component name="RecentsManager">
     <key name="CopyFile.RECENT_KEYS">
       <recent name="H:\Works\cubeExtranet\fluidbook\tools\fwstk\lib" />
@@ -123,7 +125,7 @@
       <recent name="H:\Works\cubeExtranet\fluidbook\tools\fwstk" />
     </key>
   </component>
-  <component name="RunManager" selected="Application.extract links">
+  <component name="RunManager" selected="Application.extract texts">
     <configuration default="true" type="Applet">
       <option name="POLICY_FILE" value="$APPLICATION_HOME_DIR$/bin/appletviewer.policy" />
       <method v="2">
@@ -207,6 +209,17 @@
       <option name="trusted" value="true" />
       <method />
     </configuration>
+    <configuration default="true" type="GrailsRunConfigurationType" factoryName="Grails">
+      <setting name="vmparams" value="" />
+      <setting name="cmdLine" value="run-app" />
+      <setting name="passParentEnv" value="true" />
+      <setting name="grailsApplicationRoot" value="" />
+      <setting name="launchBrowser" value="true" />
+      <setting name="launchBrowserUrl" value="" />
+      <setting name="depsClasspath" value="false" />
+      <extension name="coverage" enabled="false" merge="false" sample_coverage="true" runner="idea" />
+      <method v="2" />
+    </configuration>
     <configuration default="true" type="tests" factoryName="Attests">
       <option name="INTERPRETER_OPTIONS" value="" />
       <option name="PARENT_ENVS" value="true" />
@@ -240,8 +253,7 @@
       <option name="MODE" value="default_activity" />
       <option name="CLEAR_LOGCAT" value="false" />
       <option name="SHOW_LOGCAT_AUTOMATICALLY" value="true" />
-      <option name="SKIP_NOOP_APK_INSTALLATIONS" value="true" />
-      <option name="FORCE_STOP_RUNNING_APP" value="true" />
+      <option name="INSPECTION_WITHOUT_ACTIVITY_RESTART" value="false" />
       <option name="TARGET_SELECTION_MODE" value="SHOW_DIALOG" />
       <option name="DEBUGGER_TYPE" value="Java" />
       <Java />
@@ -267,10 +279,12 @@
       <option name="INSTRUMENTATION_RUNNER_CLASS" value="" />
       <option name="EXTRA_OPTIONS" value="" />
       <option name="INCLUDE_GRADLE_EXTRA_OPTIONS" value="true" />
+      <option name="RETENTION_ENABLED" value="No" />
+      <option name="RETENTION_MAX_SNAPSHOTS" value="2" />
+      <option name="RETENTION_COMPRESS_SNAPSHOTS" value="false" />
       <option name="CLEAR_LOGCAT" value="false" />
       <option name="SHOW_LOGCAT_AUTOMATICALLY" value="true" />
-      <option name="SKIP_NOOP_APK_INSTALLATIONS" value="true" />
-      <option name="FORCE_STOP_RUNNING_APP" value="true" />
+      <option name="INSPECTION_WITHOUT_ACTIVITY_RESTART" value="false" />
       <option name="TARGET_SELECTION_MODE" value="SHOW_DIALOG" />
       <option name="DEBUGGER_TYPE" value="Java" />
       <Java />
@@ -316,22 +330,11 @@
     <configuration name="extract texts" type="Application" factoryName="Application">
       <option name="MAIN_CLASS_NAME" value="com.fluidbook.fwstk.Main" />
       <module name="fwstk" />
-      <option name="PROGRAM_PARAMETERS" value="--input C:\Users\Vincent\Desktop\document.pdf --extractTextsMethod fluidbook --extractTexts C:\Users\Vincent\Desktop\loutres\%s%d.txt --threads 1" />
+      <option name="PROGRAM_PARAMETERS" value="--input C:\Users\vince\Desktop\original.pdf --extractTextsMethod pdfbox --extractTexts C:\Users\vince\Desktop\%s%d.txt --threads 1" />
       <method v="2">
         <option name="Make" enabled="true" />
       </method>
     </configuration>
-    <configuration default="true" type="GrailsRunConfigurationType" factoryName="Grails">
-      <setting name="vmparams" value="" />
-      <setting name="cmdLine" value="run-app" />
-      <setting name="passParentEnv" value="true" />
-      <setting name="grailsApplicationRoot" value="" />
-      <setting name="launchBrowser" value="true" />
-      <setting name="launchBrowserUrl" value="" />
-      <setting name="depsClasspath" value="false" />
-      <extension name="coverage" enabled="false" merge="false" sample_coverage="true" runner="idea" />
-      <method v="2" />
-    </configuration>
     <configuration default="true" type="JUnit" factoryName="JUnit">
       <option name="TEST_OBJECT" value="class" />
       <option name="WORKING_DIRECTORY" value="$MODULE_DIR$" />
@@ -422,32 +425,6 @@
     <supportedVersion>125</supportedVersion>
     <option name="runUnderTerminal" value="true" />
   </component>
-  <component name="SvnFileUrlMappingImpl">
-    <option name="myMappingRoots">
-      <list>
-        <SvnCopyRootSimple>
-          <option name="myVcsRoot" value="D:\Works\cubeExtranet\fluidbook\tools\fwstk" />
-          <option name="myCopyRoot" value="D:\Works\cubeExtranet\fluidbook\tools\fwstk" />
-        </SvnCopyRootSimple>
-        <SvnCopyRootSimple>
-          <option name="myVcsRoot" value="D:\Works\cubeExtranet\fluidbook\tools\fwstk" />
-          <option name="myCopyRoot" value="D:\Works\cubeExtranet\fluidbook\tools\fwstk\src\com\fluidbook\fwstk" />
-        </SvnCopyRootSimple>
-      </list>
-    </option>
-    <option name="myMoreRealMappingRoots">
-      <list>
-        <SvnCopyRootSimple>
-          <option name="myVcsRoot" value="D:\Works\cubeExtranet\fluidbook\tools\fwstk" />
-          <option name="myCopyRoot" value="D:\Works\cubeExtranet\fluidbook\tools\fwstk" />
-        </SvnCopyRootSimple>
-        <SvnCopyRootSimple>
-          <option name="myVcsRoot" value="D:\Works\cubeExtranet\fluidbook\tools\fwstk" />
-          <option name="myCopyRoot" value="D:\Works\cubeExtranet\fluidbook\tools\fwstk\src\com\fluidbook\fwstk" />
-        </SvnCopyRootSimple>
-      </list>
-    </option>
-  </component>
   <component name="TaskManager">
     <task active="true" id="Default" summary="Default task">
       <changelist id="f146bc67-2578-4de3-9db2-94d2d43e9e83" name="Default" comment="" />
@@ -492,6 +469,7 @@
       <workItem from="1606322492928" duration="1489000" />
       <workItem from="1631721143268" duration="984000" />
       <workItem from="1648579261372" duration="2928000" />
+      <workItem from="1661197921347" duration="504000" />
     </task>
     <task id="LOCAL-00001" summary="wip #1111 @0.5">
       <created>1487172253077</created>
diff --git a/resources/tools/fwstk/out/artifacts/fwstk_jar/fwstk.jar b/resources/tools/fwstk/out/artifacts/fwstk_jar/fwstk.jar
index 757f7c1..8aba199 100644
Binary files a/resources/tools/fwstk/out/artifacts/fwstk_jar/fwstk.jar and b/resources/tools/fwstk/out/artifacts/fwstk_jar/fwstk.jar differ
diff --git a/resources/tools/fwstk/src/com/fluidbook/fwstk/TextsThread.java b/resources/tools/fwstk/src/com/fluidbook/fwstk/TextsThread.java
index 24e3b99..c7e31eb 100644
--- a/resources/tools/fwstk/src/com/fluidbook/fwstk/TextsThread.java
+++ b/resources/tools/fwstk/src/com/fluidbook/fwstk/TextsThread.java
@@ -140,7 +140,7 @@ public class TextsThread extends Thread {
 					html = htmlStripper.getHTML(doc);
 					html = cleanhtml(html);
 
-					if (html.equals(memoHTML)) {
+					if (!html.equals("") && html.equals(memoHTML)) {
 						memoHTML = "";
 						continue;
 					}
@@ -154,6 +154,7 @@ public class TextsThread extends Thread {
 					String hfile = file.replaceFirst("%s", "h");
 
 					try {
+						System.out.println(":)");
 						out = new BufferedWriter(new OutputStreamWriter(
 								new FileOutputStream(pfile), "UTF8"));
 						out.write(text);
diff --git a/resources/tools/fwstk/src/com/fluidbook/fwstk/layout/Group.java b/resources/tools/fwstk/src/com/fluidbook/fwstk/layout/Group.java
index 212bf58..e35311d 100644
--- a/resources/tools/fwstk/src/com/fluidbook/fwstk/layout/Group.java
+++ b/resources/tools/fwstk/src/com/fluidbook/fwstk/layout/Group.java
@@ -4,7 +4,6 @@
  */
 package com.fluidbook.fwstk.layout;
 
-import com.sun.org.apache.xpath.internal.operations.Bool;
 import cube.util.StringUtil;
 import org.apache.pdfbox.pdmodel.common.PDRectangle;
 
diff --git a/src/PDFTools.php b/src/PDFTools.php
index fc0c1ff..0fb4f61 100644
--- a/src/PDFTools.php
+++ b/src/PDFTools.php
@@ -415,17 +415,18 @@ class PDFTools
         if ($ignoreSeparators) {
             $out .= '/sep_' . md5($ignoreSeparators);
         }
-        $out = Files::mkdir($out) . '/';
+        $out = Files::mkdir($out);
 
         $fwstk = new FWSTK();
         $fwstk->setArg('--input ' . $pdf);
         $fwstk->setArg('--extractTexts ' . $out . '%s%d.txt');
-        $fwstk->setArg('--extractTextsMethod' . $textExtraction);
+        $fwstk->setArg('--extractTextsMethod ' . $textExtraction);
         $fwstk->setArg('--threads 1');
-        if ($ignoreSeparators !== '') {
+        if ($ignoreSeparators) {
             $fwstk->setArg('--ignoreSeparators ' . $ignoreSeparators);
         }
         $fwstk->execute();
+        $fwstk->dd();
     }