Pages: « Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 Next »
Summary Period: 2003-12-03 to 2003-11-14 (Commits 2289-2338 of 3188)
Use Relation.createOnDemand() to find out if we should do a NodeManager.getNode[ByRelation]()
in Node.getProperty(). This was there in the old version of Node.getProperty() but was
kicked out in the recent rewrite, which caused NodeManager.getNode() to be called when it
shouldn't.
3 lines of code changed in:
Use getChildElement() rather than getSubnode() to get a child in HopObject.get() -
otherwise, child objects are always accessible via id, even if accessname is specified.
6 lines of code changed in:
Minor cleanup.
3 lines of code changed in:
Fix HopObject.list() which was broken after adding HopObject.list(start, length) yesterday.
(ScriptableObject.defineClass does not support method overloading)
15 lines of code changed in:
do checkNode() as first thing in HopObject.list()
4 lines of code changed in:
Fix bug where we forgot to check for NOT_FOUND in jsFunction_get()
3 lines of code changed in:
Clean up old code
3 lines of code changed in:
Always mark prototype as updated if a zip files contains some part of it,
even if it's a skin file.
6 lines of code changed in:
Prevent Nullpointer Exception in toString()
3 lines of code changed in:
Fix bug <http://helma.org/bugs/show_bug.cgi?id=304>, skins in zip files not present in skin map.
9 lines of code changed in:
Implement HopObject.list(start, length) as proposed by Tobi in
<http://helma.org/bugs/show_bug.cgi?id=303>. This is similar to
HopObject.list(), except the array it returns only contains the child
objects in the range specified by the start and length arguments.
This also does a prefetchChildren() on the specified range to optimize
database access.
27 lines of code changed in:
Cleaned up creation of SELECT statement for filter.additionalTables option.
14 lines of code changed in:
Adding dispose() as suggested by Jürg Lehni on helma-dev.
12 lines of code changed in:
Try getting a child object by ID if it isn't found as property. This is due to the
change in helma.objectmodel.db.Node.getProperty() where child nodes are
only returned as properties if either accessName or groupby is specified.
http://helma.org/bugs/show_bug.cgi?id=315
7 lines of code changed in:
Only try to get child node as property if either accessName or groupby is specified.
This fixes http://helma.org/bugs/show_bug.cgi?id=315
5 lines of code changed in:
Upgrade to Rhino snapshot from 2003/12/01
0 lines of code changed in:
Increase version to 1.3.2-pre2
3 lines of code changed in:
Include scripts directory in package build,
update version to 1.3.2-pre2
6 lines of code changed in:
Adding Linux start script from Andreas Bolka
169 lines of code changed in:
Applied fix from Juerg Lehni for parents defined as COMPLEX_REFERENCE that
was posted on September 19 and somehow fell through the gutter.
4 lines of code changed in:
Set application classloader on the Rhino context to our app classloader.
7 lines of code changed in:
Sorry Stefan, I was completely exhausted ;-)
10 lines of code changed in:
add commons-logging.jar to classpath
3 lines of code changed in:
make sure logdir is not null in getLogger()
3 lines of code changed in:
Checking in Commons Logging support from Daniel, plus a major rewrite
of the Helma logging framework.
794 lines of code changed in:
Various minor enhancements and cleanups:
* Implemented rootId property to set the id of the app's root object
* Do not call invokeFunction with a null argument array
* Implement toString() method
* Plus some more minor stuff
20 lines of code changed in:
Changed logging in CacheMap to be Logging framework independent
(just use the application's logEvent() method)
14 lines of code changed in:
Added HopObject constructor, improved variable naming in jsConstructor
26 lines of code changed in:
Adding Jakarta Commons Logging 1.0.3 jar
0 lines of code changed in:
Avoid unnecessary log messages.
3 lines of code changed in:
* Make check for _prototype and _name column case insensitive when inserting relational objects
* Implemented feature to evict certain prototypes from cache when they are received via replication.
To activate this, add the line
_evictOnReplication = true
to the type.property file
27 lines of code changed in:
Slightly improved error message when DBSource could not be loaded
(include possibility that driver couldn't be loaded)
3 lines of code changed in:
Rename pushStringBuffer() and popStringBuffer() to push() and pop().
Keeping old versions for backwards-compatibility for the time being.
25 lines of code changed in:
Unwrap arguments to Java objects in addPart(). Also, if the part is a java.io.File,
set the part's file name if no explicit file name is given.
12 lines of code changed in:
Rethrow wrapped ConcurrencyExceptions. Also updated code to
rethrow wrapped TimeoutExceptions.
13 lines of code changed in:
Add leading space in " WHERE " when composing select statements.
Fixes bug http://www.helma.org/bugs/show_bug.cgi?id=313
5 lines of code changed in:
Reverse sequence of <get-from-nodemgr> and <get-from-propmap> in
getProperty(). This may return old cached properties from the propmap when it
shouldn't, but it fixes the much worse bug where embedded-db-collections are
created from scratch again and again.
21 lines of code changed in:
minor formatting fix
3 lines of code changed in:
Remove code that was commented out and isn't coming back
2 lines of code changed in:
Re-add check if Relation is PRIMITIVE or REFERENCE when creating properties in createNode
(was removed yesterday, not sure it's really needed)
Remove code that was commented out and isn't coming back
4 lines of code changed in:
Set initial TypeInfo.lastUpdate to -1 so a Prototype with no files (lastUpdate == 0)
will be evaluated at least once. This fixes a bug with empty prototypes not having their
prototype set (i.e. get(), list(), href() etc were unavailable).
13 lines of code changed in:
Add an intermediate step when fetching properties from result sets in createNode() so that
we can be sure to use the proper DbMapping for column-to-property mappings (which may change
according to the object's prototype column)
This should fix bug 310. http://helma.org/bugs/show_bug.cgi?id=310
33 lines of code changed in:
Implement protected setName() method to set the property's name
11 lines of code changed in:
Always return our own DbColumns[]. This is needed in order to add/change mappings
in extended prototypes.
5 lines of code changed in:
Reworked Mail object contstuctor code:
* Do not set smtp host unless it is explicitely set in the properties
* Set message encoding as defined in mail.charset property, defaults to ISO8859-15
* Cache Mail session in Mail object prototype
* Store app properties as Java fields in prototype rather than as JS property in constructor
63 lines of code changed in:
Use Session.getInstance() rather than Session.getDefaultInstance(). This creates a new Session for
each Mail() object, creating an additional overhead of ~10millis, but it picks up changes in the
properties files, fixing bug 308.
http://www.helma.org/bugs/show_bug.cgi?id=308
3 lines of code changed in:
Update to JavaMail 1.3.1 FCS
0 lines of code changed in:
Remove unused imports
4 lines of code changed in:
Added experimental res.forward() to forward a request to a local (static) resource.
The base resource must be mounted as <app>.protectedStatic in the apps.properties file.
Currently no Content-Type is set on the response.
47 lines of code changed in:
Added experimental res.forward() to forward a request to a local (static) resource.
The base resource must be mounted as <app>.protectedStatic in the apps.properties file.
39 lines of code changed in:
Pages: « Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 Next »
Generated by StatCVS 0.2.2