Merge lp:~zorba-coders/zorba/fread-pdf-trunk into lp:zorba

[Cezar Andrei]

Add read-pdf module, which reads the text from a pdf doc and renders its pages to images.
Add java.library.path to the jvm in util-jvm module.
Make doc comments for createBaser64Binary more explicit on what parameters they expect and what they do.
Change, return value to xs_int for getIntValue() method.

Make doc comments for createBaser64Binary more explicit on what parameters they expect and what they do.
Change, return value to xs_int for getIntValue() method.

[Chris Hillery (ceejatec)]

You changed the URIs of the test modules back to www.zorba-xquery.com, instead of zorba-tests.28msec.us. That will cause test failures. Make sure you've merged from the trunk recently.

review: Needs Fixing

[Chris Hillery (ceejatec)]

(Other than that the changes look fine)

[Cezar Andrei (cezar-andrei)]

It's up to date now.

review: Approve

[Chris Hillery (ceejatec)]

You still haven't corrected the URIs of the test modules, though (test/rbkt/modules/CMakeLists.txt).

[Cezar Andrei (cezar-andrei)]

Let me add the new module.

review: Needs Fixing

[Cezar Andrei (cezar-andrei)]

The ref to the new module is in, please review it too: lp:zorba/read-pdf-module at
https://code.launchpad.net/~zorba-coders/zorba/fread-pdf-the-module .

[Cezar Andrei (cezar-andrei)]

Back to zorba-tests.28msec.us in test/rbkt/modules/CMakeLists.txt and mention in Changelog.

[Chris Hillery (ceejatec)]

Great, thanks.

review: Approve

[Zorba Build Bot (zorba-buildbot)]

Attempt to merge into lp:zorba failed due to conflicts:

text conflict in ChangeLog

[Matthias Brantner (matthias-brantner)]

It looks nice (without being able to testing it ;-)

- The module should follow XQuery coding conventions. For example, no camel-case
but function and variable names using dashes.

- The read-pdf:renderToImages-internal function is not documented. This will probably cause
the xqdoc test to fail.

- There is a conflict in the ChangeLog.

- The schema contains some commented out code => remove?

- The declaration of the $options variable for the -internal functions should not allow for
an optional occurrence indicator.

review: Needs Fixing

[Cezar Andrei (cezar-andrei)]

I fixed all the comments with the exception of the optional occurrence indicator for internal functions. I tested it with empty sequence and works fine.

[Zorba Build Bot (zorba-buildbot)]

Validation queue starting for merge proposal.
Log at: http://zorbatest.lambda.nu:8080/remotequeue/fread-pdf-trunk-2012-09-20T03-03-43.261Z/log.html

[Zorba Build Bot (zorba-buildbot)]

Validation queue job fread-pdf-trunk-2012-09-20T03-03-43.261Z is finished. The final status was:

All tests succeeded!

[Zorba Build Bot (zorba-buildbot)]

Voting does not meet specified criteria. Required: Approve > 1, Disapprove < 1, Needs Fixing < 1, Pending < 1. Got: 1 Approve, 2 Needs Fixing.

[Paul J. Lucas (paul-lucas)]

It doesn't look like the fpdf.* files were added via "bzr add" -- were they supposed to be?

[Cezar Andrei (cezar-andrei)]

Paul,

If you're asking about my last commit msg, it is as it should be just one line change, by deleting the tag name, it makes trunk use the latest HEAD version of the module branch. The files in util-jvm were already reviewed and merged see this merge: https://code.launchpad.net/~zorba-coders/zorba/fread-pdf-utiljvm-module/+merge/125332 .

The changes in file: modules/ExternalModules.conf'

117 --- modules/ExternalModules.conf 2012-09-17 00:36:37 +0000
118 +++ modules/ExternalModules.conf 2012-09-21 15:25:26 +0000
119 @@ -39,11 +39,12 @@
123 +read-pdf bzr lp:zorba/read-pdf-module
128 -util-jvm bzr lp:zorba/util-jvm-module zorba-2.6
129 +util-jvm bzr lp:zorba/util-jvm-module

[Matthias Brantner (matthias-brantner)]

The module works pretty decent. I was able to extract text or generate images for several pdfs without any problems.
There are some minor things that should be discussed and/or fixed:

- the error seems to be too general, essentially it always raises JAVA-EXCEPTION no matter what goes wrong (e.g. it the given input is not a valid pdf)

- the java stack trace seems to be sent to standard error

- Renders the each page of the PDF document as an image. => Renders each page of the PDF document as an image.

- the names of the private functions should also adhere to the code conventions renderToImages => render-to-images

- make xqdoc failes because the comments seem to contain invalid xml
</home/mbrantner/zorba/build/URI_PATH/com/zorba-xquery/www/modules/project_xqdoc.xq>:142,9: user-defined error [err:UE004]: Error processing module zerr:ZXQD0002 - " This module provides funtionality to read the text from PDF documents and
 to render PDF documents to images.
 <a href="http://pdfbox.apache.org">Apache PDFBox</a> library is used to
 implement these functions.
 <br />
 <br />
 <b>Note:</b> Since this module has a Java library dependency a JVM required
 to be installed on the system. For Windows: jvm.dll is required on the system
 path ( usually located in "C:\Program Files\Java\jre6\bin\client".
 <b>Note:<b> For Debian based Linux distributions install PdfBox and FontBox
 packages: sudo apt-get install libpdfbox-java libfontbox-java
": can not parse as XML for xqdoc: loader parsing error: Opening and ending tag mismatch: b line 0 and root
; raised at /home/mbrantner/zorba/sandbox/src/runtime/errors_and_diagnostics/errors_and_diagnostics_impl.cpp:81

- adapt the year in "Copyright 2006-2009 The FLWOR Foundation." in the .xq file (and some other files also)

- would it make sense to return one string per page in the pdf instead of one big string?

- remove commented out code in read-pdf.cpp

- valgrind shows tons of invalid writes. Why? Are they critical? Is there anything we can do?

- would it make sense to return the images in a streaming fashion (i.e. don't create all base64's in a vector)?

- encoding each image shouldn't be necessary and will probably we wasted effort because the images might be written to a file in their binary form

review: Needs Fixing

[Cezar Andrei (cezar-andrei)]

See answers inline:
- the error seems to be too general, essentially it always raises JAVA-EXCEPTION no matter what goes wrong (e.g. it the given input is not a valid pdf)
I adapted the error msg to be more clear/specific.

- the java stack trace seems to be sent to standard error
Goes to std err.

- Renders the each page of the PDF document as an image. => Renders each page of the PDF document as an image.
Done.

- the names of the private functions should also adhere to the code conventions renderToImages => render-to-images
Done.

- make xqdoc failes because the comments seem to contain invalid xml
</home/mbrantner/zorba/build/URI_PATH/com/zorba-xquery/www/modules/project_xqdoc.xq>:142,9: user-defined error [err:UE004]: Error processing module zerr:ZXQD0002 - " This module provides funtionality to read the text from PDF documents and
 to render PDF documents to images.
 <a href="http://pdfbox.apache.org">Apache PDFBox</a> library is used to
 implement these functions.
 <br />
 <br />
 <b>Note:</b> Since this module has a Java library dependency a JVM required
 to be installed on the system. For Windows: jvm.dll is required on the system
 path ( usually located in "C:\Program Files\Java\jre6\bin\client".
 <b>Note:<b> For Debian based Linux distributions install PdfBox and FontBox
 packages: sudo apt-get install libpdfbox-java libfontbox-java
": can not parse as XML for xqdoc: loader parsing error: Opening and ending tag mismatch: b line 0 and root
; raised at /home/mbrantner/zorba/sandbox/src/runtime/errors_and_diagnostics/errors_and_diagnostics_impl.cpp:81
Done.

- adapt the year in "Copyright 2006-2009 The FLWOR Foundation." in the .xq file (and some other files also)
Done.

- would it make sense to return one string per page in the pdf instead of one big string?
The API doesn't alow it, but I added two more optional options, to insert a user defined string at the start and end of each page.

- remove commented out code in read-pdf.cpp
Done.

- valgrind shows tons of invalid writes. Why? Are they critical? Is there anything we can do?
Jvm always shows in valgrind, even if nothing is done with it. I was careful to remove any allocated memory.

- would it make sense to return the images in a streaming fashion (i.e. don't create all base64's in a vector)?
No, because it's a push write of all images. And as discussed, optimize only a copy in some cases isn't worth the effort.

- encoding each image shouldn't be necessary and will probably we wasted effort because the images might be written to a file in their binary form
Done.

[Cezar Andrei (cezar-andrei)]

review: Approve

[Zorba Build Bot (zorba-buildbot)]

Validation queue starting for merge proposal.
Log at: http://zorbatest.lambda.nu:8080/remotequeue/fread-pdf-trunk-2012-09-22T21-29-56.025Z/log.html

[Zorba Build Bot (zorba-buildbot)]

Validation queue job fread-pdf-trunk-2012-09-22T21-29-56.025Z is finished. The final status was:

All tests succeeded!

[Zorba Build Bot (zorba-buildbot)]

Voting does not meet specified criteria. Required: Approve > 1, Disapprove < 1, Needs Fixing < 1, Pending < 1. Got: 1 Approve, 2 Pending.

[Matthias Brantner (matthias-brantner)]

review: Approve

[Zorba Build Bot (zorba-buildbot)]

Validation queue starting for merge proposal.
Log at: http://zorbatest.lambda.nu:8080/remotequeue/fread-pdf-trunk-2012-09-25T20-57-55.26Z/log.html

[Zorba Build Bot (zorba-buildbot)]

Validation queue job fread-pdf-trunk-2012-09-25T20-57-55.26Z is finished. The final status was:

All tests succeeded!

[Zorba Build Bot (zorba-buildbot)]

Voting does not meet specified criteria. Required: Approve > 1, Disapprove < 1, Needs Fixing < 1, Pending < 1. Got: 2 Approve, 1 Pending.

[Chris Hillery (ceejatec)]

Cezar - The remote queue is still not finding the jars, although they are installed correctly in /usr/share/java as requested. (This is also probably why the experimental build isn't including the module.) The problem, I believe, is that you are using FIND_FILE() to find them, and you only list a few specific version numbers to find - I installed 1.7.1 from the URL you sent, but that isn't one of the versions listed in CMakeLists.txt in the read-pdf module.

The correct solution is to use ZORBA_FIND_JAR(), which will look for .jar files with any embedded version number. See src/com/zorba-xquery/www/modules/CMakeLists.txt in the data-formatting module for an example. If you make that change in the module, this merge proposal should succeed.

[Chris Hillery (ceejatec)]

I have re-assigned my review request to Juan, since I will not be around much longer to vote on this proposal.

Juan, all you need to review is the CMake stuff in the lp:zorba/read-pdf-module branch, specifically to ensure that it uses FIND_ZORBA_JAR() appropriately. Once Cezar makes that change, you can vote Approve on this proposal.

[Cezar Andrei (cezar-andrei)]

I've modified read-pdf's CMakeLists.txt to use ZORBA_FIND_JAR instead of FIND_FILE.

review: Approve

[Zorba Build Bot (zorba-buildbot)]

Validation queue starting for merge proposal.
Log at: http://zorbatest.lambda.nu:8080/remotequeue/fread-pdf-trunk-2012-09-25T23-51-47.915Z/log.html

[Zorba Build Bot (zorba-buildbot)]

The attempt to merge lp:~zorba-coders/zorba/fread-pdf-trunk into lp:zorba failed. Below is the output from the failed tests.

CMake Error at /home/ceej/zo/testing/zorbatest/tester/TarmacLander.cmake:274 (message):
  Validation queue job fread-pdf-trunk-2012-09-25T23-51-47.915Z is finished.
  The final status was:

  No tests were run - build or configure step must have failed.

  Not commiting changes.

Error in read script: /home/ceej/zo/testing/zorbatest/tester/TarmacLander.cmake

[Zorba Build Bot (zorba-buildbot)]

Validation queue starting for merge proposal.
Log at: http://zorbatest.lambda.nu:8080/remotequeue/fread-pdf-trunk-2012-09-26T04-11-47.786Z/log.html

[Chris Hillery (ceejatec)]

The read-pdf module was found and built with this latest RQ run.

However, it has no tests. I personally don't think that's acceptable, even for a beta release.

[Zorba Build Bot (zorba-buildbot)]

The attempt to merge lp:~zorba-coders/zorba/fread-pdf-trunk into lp:zorba failed. Below is the output from the failed tests.

CMake Error at /home/ceej/zo/testing/zorbatest/tester/TarmacLander.cmake:274 (message):
  Validation queue job fread-pdf-trunk-2012-09-26T04-11-47.786Z is finished.
  The final status was:

  1 tests did not succeed - changes not commited.

Error in read script: /home/ceej/zo/testing/zorbatest/tester/TarmacLander.cmake

[Chris Hillery (ceejatec)]

Error message from failed test:

</home/ceej/zo/testing/zorbatest/tester/remotequeue/fread-pdf-trunk-2012-09-26T04-11-47.786Z/bzr/zorba/build/URI_PATH/com/zorba-xquery/www/modules/read-pdf_0.xq>:127,18: Zorba error [zerr:ZXQP0008]: "{http://www.zorba-xquery.com/modules/read-pdf}extract-text-internal#2": function implementation not found

Remember that the RQ runs in release mode.

[Cezar Andrei (cezar-andrei)]

I added texts for all functions.

review: Approve

[Zorba Build Bot (zorba-buildbot)]

Attempt to merge into lp:zorba failed due to conflicts:

text conflict in ChangeLog

[lp:~zorba-coders/zorba/fread-pdf-trunk]

10955. By Cezar Andrei <email address hidden> on 2012-09-26

Merge from trunk, resolved ChangeLog conflict.

10956. By Cezar Andrei <email address hidden> on 2012-09-27

Merge from trunk.
Resolved changeLog conflict.

10957. By Cezar Andrei <email address hidden> on 2012-09-27

Merge from trunk. Resolved conflict in ExternalModules.conf.

10958. By Cezar Andrei <email address hidden> on 2012-09-27

Add tag to read-pdf module.

10959. By Cezar Andrei <email address hidden> on 2012-10-02

Merge from trunk, resolve Changelog conflict.

Unmerged revisions

=== modified file 'ChangeLog'
--- ChangeLog	2012-09-26 21:29:08 +0000
+++ ChangeLog	2012-09-26 23:06:48 +0000
@@ -17,7 +17,8 @@
   * Implemented semantics of null for comparison and arithmetics operations.
   * Positional pagination support for index probes 
   * Recognize the no-copy pragma to avoid copying nodes before insertion into a collection.
-	* Added createUntypedAtomic to API's ItemFactory.
+  * Added createUntypedAtomic to API's ItemFactory.
+  * Adding new external module read-pdf, it converts PDF documents to text or rendered images.
 
 Optimizations:
   * New memory management for compiler expressions (no more ref counting)

=== modified file 'cmake_modules/ZorbaModule.cmake'
--- cmake_modules/ZorbaModule.cmake	2012-09-17 00:36:37 +0000
+++ cmake_modules/ZorbaModule.cmake	2012-09-26 23:06:48 +0000
@@ -114,9 +114,20 @@
 #              relative to CMAKE_CURRENT_SOURCE_DIR)
 #       LINK_LIBRARIES - (optional) List of libraries to link external
 #              function library against
+#       CONFIG_FILES - (optional) List of files to configure with package
+#              information; see below
 #       TEST_ONLY - (optional) Module is for testcases only and should not
 #              be installed
 #
+# CONFIG_FILES - any files specific here will be copied to
+# CMAKE_CURRENT_BINARY_DIR using CONFIGURE_FILE(). They may contain
+# the following @VARIABLES@ which will be substituted:
+#       ZORBA_MODULE_RELATIVE_DIR - directory portion of mangled URI
+#       ZORBA_MODULE_LIBFILE_WE - filename (without extension) portion of
+#              mangled URI
+# The input files should have a .in extension. The resulting file in
+# the build directory will have the .in removed.
+#
 # QQQ this currently doesn't support modules with multiple component
 # .xq files. (Neither does Zorba's automatic loading mechanism, so
 # this probably isn't a huge deal, but worth thinking about.)
@@ -125,7 +136,7 @@
 # file enough to deduce the URI and version?
 MACRO (DECLARE_ZORBA_MODULE)
   # Parse and validate arguments
-  PARSE_ARGUMENTS(MODULE "LINK_LIBRARIES;EXTRA_SOURCES"
+  PARSE_ARGUMENTS(MODULE "LINK_LIBRARIES;EXTRA_SOURCES;CONFIG_FILES"
     "URI;FILE;VERSION" "TEST_ONLY" ${ARGN})
   IF (NOT MODULE_FILE)
     MESSAGE (FATAL_ERROR "'FILE' argument is required for ZORBA_DECLARE_MODULE()")
@@ -353,6 +364,20 @@
       "${version_infix}" "" 1 "${MODULE_TEST_ONLY}")
   ENDFOREACH (version_infix)
 
+  # Configure any module-specified config files.
+  SET (ZORBA_MODULE_RELATIVE_DIR ${module_path})
+  SET (ZORBA_MODULE_LIBFILE_WE ${module_filewe})
+  FOREACH (_config_file ${MODULE_CONFIG_FILES})
+    # Strip off .in - can't use GET_FILENAME_COMPONENT as it always removes
+    # the longest possible extension
+    STRING (REGEX REPLACE "\\.in$" "" _config_filename_we "${_config_file}")
+    IF (NOT IS_ABSOLUTE "${_config_file}")
+      SET (_config_file "${CMAKE_CURRENT_SOURCE_DIR}/${_config_file}")
+    ENDIF (NOT IS_ABSOLUTE "${_config_file}")
+    CONFIGURE_FILE (${_config_file}
+      "${CMAKE_CURRENT_BINARY_DIR}/${_config_filename_we}" @ONLY)
+  ENDFOREACH (_config_file)
+
   # Last but not least, whip up a test case that ensures the module
   # can at least be compiled. Don't bother for test-only modules
   # (presumably they're there to be tested!).

=== modified file 'include/zorba/item_factory.h'
--- include/zorba/item_factory.h	2012-09-26 17:09:11 +0000
+++ include/zorba/item_factory.h	2012-09-26 23:06:48 +0000
@@ -123,8 +123,8 @@
       /** \brief Creates a Base64Binary Item
        *         see [http://www.w3.org/TR/xmlschema-2/#base64Binary]
        *
-       * @param aBinData a pointer to the base6c4 binary data.
-       * @param aLength the length of the base64 binary data.
+       * @param aBinData a pointer to the base64 encoded data. The data is copied from aBinData.
+       * @param aLength the length of the base64 encoded data.
        * @return The Base64Binary Item.
        */
       virtual Item
@@ -133,7 +133,7 @@
       /** \brief Creates a Base64Binary Item
        *         see [http://www.w3.org/TR/xmlschema-2/#base64Binary]
        *
-       * @param aStream A stream containing the Base64 encoded data.
+       * @param aStream A stream containing the Base64 encoded data. The data is copied from aStream imediately.
        * @return the Base64Binary Item.
        */
       virtual Item
@@ -142,11 +142,11 @@
       /** \brief Creates a Base64Binary Item
        *         see [http://www.w3.org/TR/xmlschema-2/#base64Binary]
        *
-       * @param aBinData the data in binary form. The data is copied from aBinData.
-       * @param aLength the length of the data
+       * @param aBinData the data in binary form (not encoded). The data is copied from aBinData.
+       * @param aLength the length of the binary data
        * @return the Base64Binary Item.
        */
-      virtual Item 
+      virtual Item
       createBase64Binary(const unsigned char* aBinData, size_t aLength) = 0;
 
       /** \brief Creates a streamable Base64Binary Item
@@ -742,7 +742,7 @@
        * @param aNames A vector containing the name and value of each pair.
        */
       virtual Item createJSONObject(std::vector<std::pair<Item, Item> >& aNames) = 0;
-      
+
       /**
        * Create a JSON Array containing the specified items.
        *

=== modified file 'modules/ExternalModules.conf'
--- modules/ExternalModules.conf	2012-09-17 00:36:37 +0000
+++ modules/ExternalModules.conf	2012-09-26 23:06:48 +0000
@@ -39,11 +39,12 @@
 languages       bzr  lp:zorba/languages-module        zorba-2.6
 oauth           bzr  lp:zorba/oauth-module            zorba-2.6
 process         bzr  lp:zorba/process-module          zorba-2.6
+read-pdf        bzr  lp:zorba/read-pdf-module
 security        bzr  lp:zorba/security-module         zorba-2.6
 system          bzr  lp:zorba/system-module           zorba-2.6
 xqxq            bzr  lp:zorba/xqxq-module             zorba-2.6
 email           bzr  lp:zorba/email-module            zorba-2.6
-util-jvm        bzr  lp:zorba/util-jvm-module         zorba-2.6
+util-jvm        bzr  lp:zorba/util-jvm-module         
 schema-tools    bzr  lp:zorba/schema-tools-module     zorba-2.6
 stack           bzr  lp:zorba/stack-module            zorba-2.6
 queue           bzr  lp:zorba/queue-module            zorba-2.6

=== modified file 'src/store/naive/atomic_items.h'
--- src/store/naive/atomic_items.h	2012-09-17 00:36:37 +0000
+++ src/store/naive/atomic_items.h	2012-09-26 23:06:48 +0000
@@ -1776,7 +1776,7 @@
   friend class AtomicItem;
 
 protected:
-  int32_t theValue;
+  xs_int theValue;
 
 protected:
   IntItem(xs_int aValue) : theValue(aValue) {}
@@ -1792,7 +1792,7 @@
 
   xs_long getLongValue() const { return static_cast<xs_long>(theValue); }
 
-  int32_t getIntValue() const { return theValue; }
+  xs_int getIntValue() const { return theValue; }
 
   store::SchemaTypeCode getTypeCode() const { return store::XS_INT; }
 

=== modified file 'test/rbkt/modules/CMakeLists.txt'
--- test/rbkt/modules/CMakeLists.txt	2012-09-17 00:36:37 +0000
+++ test/rbkt/modules/CMakeLists.txt	2012-09-26 23:06:48 +0000
@@ -27,8 +27,9 @@
 DECLARE_ZORBA_MODULE(URI "http://zorba-tests.28msec.us/modules/B" VERSION 1.0
   FILE "${CMAKE_CURRENT_SOURCE_DIR}/module-B.xq" TEST_ONLY)
 
+# This one also has a CONFIG_FILE
 DECLARE_ZORBA_MODULE(URI "http://zorba-tests.28msec.us/modules/ext"
-  VERSION 2.0 FILE "ext2.xq" TEST_ONLY)
+  VERSION 2.0 FILE "ext2.xq" TEST_ONLY CONFIG_FILES ext2_config.txt.in)
 DECLARE_ZORBA_MODULE(URI "http://zorba-tests.28msec.us/modules/ext"
   VERSION 1.0 FILE "ext.xq" TEST_ONLY)
 

=== added file 'test/rbkt/modules/ext2_config.txt.in'
--- test/rbkt/modules/ext2_config.txt.in	1970-01-01 00:00:00 +0000
+++ test/rbkt/modules/ext2_config.txt.in	2012-09-26 23:06:48 +0000
@@ -0,0 +1,1 @@
+This is a test: "@ZORBA_MODULE_RELATIVE_DIR@" "@ZORBA_MODULE_LIBFILE_WE@"