Just released 0.9.0.7 alpha. It contains only a single fix as I realized, that the file list did not get updated after an index switch. The issue was caused by exceptions during the update and serialization of user options.
Now the file and file type lists get updated properly and file search works as expected.
Unfortunately an annoying bug has creeped into the latest release. While refactoring the command line parsing I did not test it with relative file paths, and now I realized, that they do not work, resulting in a somewhat cryptic message when starting either the indexing or the search tool, that the operation is not supported for relative URIs ( namely retrieving the absolute path ).
The issue has been fixed now and I've uploaded a new version of the 0.9.0.6 package in the files section and also updated the source code of the trunk and the 0.9.0.6 branch.... read more
I've been busy the last couple of months and had not much motivation and time to continue. Thus this release is only a small one and resulted from personal needs while using CodeXCavator at work.
It's now possible to pass index configuration file lists or directories as arguments to the indexing and search tools. This facilitates some things when dealing with many indexes and also circumvents the command line length limitation.... read more
A happy new year to everyone!
As i noticed, that keyboard navigation between matches, only works for matches returned by global search but not for in-file search, i've updated 0.9.0.5 and added match navigation for in-file search matches. You can either use the navigational buttons on the in-file search pane or the keyboard short cuts Ctrl+Up ( previous match ) and Ctrl+Down ( next match ).
I wish you a merry christmas and bring a present with me: Version 0.9.0.5 alpha of CodeXCavator has been released!
Beside minor fixes, I've finally integrated in-file multi-search. It can be triggered using the search button in the lower right corner of the text viewer or by using the keyboard short cut Ctrl + F. Details about this functionality can be found in the Wiki.
Please refer to the readme.txt in the files section for further information on other changes made in this release.
No update until now, as I had no time, was on holidays, and recently had no motivation. Will start on in-file search as soon as possible.
The previously announced features "Tabbed search" and "Multi-search within the file viewer" will be addressed in the next milestone due to lack of time.
A new version 0.9.0.4 alpha of CodeXCavator has just been released. It mostly includes some minor fixes and a bunch of useability improvements. Especially the scrolling and handling within the search results list has been improved, such that you now can expand all search hit occurrences within a single file and scroll through them without any problems. You can also jump to the occurrence within the file, simply by opening it.
Furthermore you can now toggle the live search functionality individually for each search text box.... read more
After improving text browsing performance in milestome 4.0 the next milestone 5.0 will be used to improve the overall useability of the search tool. Many
useful suggestions have been made by colleagues and will be addressed during that phase.
Besides some minor improvements, I'll try to add two in my opinion useful features:
Just released version 0.9.0.3 alpha of the CodeXCavator. One major improvement is the replacement of the FlowDocumentViewer used for displaying syntax highlighted source code within the file browser by a custom TextViewerControl user control. This improves source code browsing performance, especially when opening large source code files.
Furthermore some stability issues when using the indexer on badly configured indexes have been fixed. The indexer now reports errors, which occur during indexing in a graphical way. ... read more
I've almost finished implementing the new text browser control. It now supports selection of text via mouse, scrolling via keys and text hit events, which make context menu implementation and text based navigation easier.
Full line selection when clicking on the line numbers column is not working yet and changing the mouse cursor when hovering over text, must be implemented in a better and more generic way.... read more
As the current file viewer suffers from huge delays, before the opened file is actually shown ( up to several seconds ), i've tried to identify the bottleneck, and the bottleneck was WPF's flow document viewer control I've used and because i've tried to prepare the whole document in advance including all highlighting.
Thus I've started writing a new text viewer control on Sunday and it's progressing well. I already can display and scroll syntax highlighted text including colored background highlighting of matches. ... read more
Unfortunately I did not notice until now, that the published binary packages did not work due to missing app-config files. I assumed that they would be embedded with the executables automatically, but they did not. Thus the tools failed to work, as the Lucene assemblies could not have been loaded because they are located in the Externals folder. For this case i've added probing paths to the app-config file, but without the config file being in the binary package, this would of course not work. ... read more
Currently the file viewer suffers from bad performance. Opening a large source file with much syntax highlighting takes up too much time ( several seconds ). In the current version the viewer uses a flow document to display the source code and syntax highlighting is performed in advance while creating the flow document.
I also tried to write a simple text viewer control based on FormattedText. However the performance is still bad. I made some measurements and it turns out that most of the time, is taken by configuring the FormattedText object (i.e. setting text color for different portions of text), measuring it and rendering it.... read more
Released an update few days ago. This update includes a silent mode and an auto close option for the indexing tool, such that it can be run from a batch script or a scheduled task without the necessity of user interaction.
Furthermore the update includes a PluginManager component, which dynamically loads assemblies from a "plugin" folder which must be placed in the same directory as the CodeXCavator.Enigne assembly. The assemblies are scanned for certain types and those types are registered within the the enigne, such that they can be used.... read more
Today the initial alpha version of the CodeXCavator tool has been released. You can download it from the files section. It includes index creation and index search functionality. It also already supports code tagging. It has some performance issues, when opening large source code files.