mirrors/zotero
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages 1 Wiki Activity Actions
604 commits 9 branches 366 tags 311 MiB
Commit graph

299 commits

Author SHA1 Message Date
Simon Kornblith
260ce80086 - Search results scraping for TLC. This is the last of the library scrapers. 
- Minor fixes to ingester utilities.
 2006-06-24 15:38:53 +00:00
Dan Stillman
97940c7470 Replaced all instances of "Firefox Scholar" (not counting the repository URL) with "Scholar for Firefox" for now 2006-06-24 09:08:12 +00:00
Dan Stillman
720960feb9 Addresses #5, Add as many item types as possible 
New item types from Elena
 2006-06-24 08:28:37 +00:00
Dan Stillman
42578ace59 Fixes #61, creator caching sometimes malfunctions 2006-06-23 22:00:39 +00:00
Simon Kornblith
2a74e88416 - Make generalized function for finding search results case insensitive 
- Scrape DRA search results
 2006-06-23 20:09:48 +00:00
Simon Kornblith
098078627c - Make events listening for DOMContentLoaded listen for load, because DOMContentLoaded does not seem ready for prime time (hey, it's undocumented, what can you expect) 
- Make Amazon scraper work with multiple documents
- Fix bugs in processDocuments
- Make Scholar.Ingester.Utilities.getItemArray() willing to take an array of DOM nodes to search for links, and finally take advantage of the fact that objects have no length
 2006-06-23 03:02:30 +00:00
Simon Kornblith
470f7c463f The Voyager scraper now actually works on the search results page. 2006-06-22 20:50:57 +00:00
Simon Kornblith
3890e5f122 - Made ingester automatically create hidden browser objects, given a window object. This should make things much easier for both David and me. 
- Multiple item detection code is now a part of the scraperJavaScript, rather than the scrapeDetectCode, and code to choose which items to add is part of Scholar.Ingester.Utilities, accessible from inside scrapers. The alternative approach would result in one request (or, in the case of JSTOR, three requests) per new item, while in some cases (e.g. Voyager) only one request is necessary to get all of the items.
 2006-06-22 15:50:46 +00:00
Dan Stillman
726364d091 Scholar.History -- i.e. undo/redo functionality 
Partially integrated into data layer, but I'm waiting to commit that part until I'm sure it won't break everything
 2006-06-22 14:01:54 +00:00
Simon Kornblith
ca3a0e6e5d Beginnings of search result scraping (does not yet actually do the scraping, but does present the menu) 2006-06-22 02:43:40 +00:00
Dan Stillman
6b2d9bb3e6 Don't break the extension right before we show it 2006-06-21 16:06:41 +00:00
Dan Stillman
336f92a832 Set repository check timer to 1 hour during development 2006-06-21 16:04:45 +00:00
Simon Kornblith
6d1e447154 - Remove load eventListener after it has been called once 
- Capture editors from Google Books
 2006-06-21 15:18:18 +00:00
Simon Kornblith
7d3deb5b9f - Make Scholar.Ingester.Utilities.loadDocument() attach an event handler to load rather than DOMContentLoaded to resolve an issue with the Ex Libris/Aleph scraper (VCU) 
- When possible, corporate creators/contributors are categorized with their own RDF types (prefixDummy + "corporateCreator/corporateContributor)
- Remove extraneous debug code in extensions
 2006-06-21 01:41:07 +00:00
Dan Stillman
6c89acbe0d Scholar.inArray(needle, haystack) and Scholar.arraySearch(needle, haystack) -- versions of the PHP functions for JS 2006-06-20 17:32:40 +00:00
Simon Kornblith
09d79d6dd7 Fix overly optimistic JSTOR scraper 2006-06-20 17:06:41 +00:00
Dan Stillman
a3df0c39e2 - DB parameters can now be bound using the native JS type rather than by specifying the type explicitly (e.g. Scholar.DB.query(sql, [1, 2, "hello"]) -- for use when the data is generated internally and trusted, obviously 
- Don't try to display an SQLite error when it's "not an error" (i.e. when the error is in something else)

- Switch to nsIFile instead of nsILocalFile to retrieve the profile directory
 2006-06-20 15:42:01 +00:00
Dan Stillman
d239e6e4ba Typo in trigger() call in Collection.addItem() (thanks David) 2006-06-20 15:28:20 +00:00
Dan Stillman
95ca76545f QA testing of extension performance aboard trains 
Fix in Collection.erase() -- when the DB methods started returning values in their native type, the collection id became an int rather than a string and "new Array(this._id)" became a length declaration rather than an elements declaration
 2006-06-20 15:23:44 +00:00
Simon Kornblith
20369f41b3 - Move commonly used scraper functions to ingester.js, rather than re-defining them in each scraper. This breaks Piggy Bank compatibility in our scrapers, but we will still be able to export our scrapers in a Piggy Bank compatible form. 
- Better handling of scraper RDF to item mapping.
- Improved date handling. All scrapers now return ISO-style dates when possible.
 2006-06-18 19:04:32 +00:00
Simon Kornblith
3d881eec13 - Make scrapers return standard ISO-style YYYY-MM-DD dates. Still need to work on journal article scrapers. 
- Ingester lets callback function save items, rather than saving them itself.
- Better handling of multiple items in API, although no scrapers currently implement this.
 2006-06-17 21:21:15 +00:00
Dan Stillman
32ce0da44a Restore the statement.reset() in valueQuery() that I accidentally deleted in r216 2006-06-16 22:17:46 +00:00
Dan Stillman
bfa25cae37 Display SQLite error in exceptions thrown from commitTransaction() and rollbackTransaction() (like the ones on field value updates that I'm trying to figure out at the moment) 2006-06-16 22:15:07 +00:00
Dan Stillman
e8ed4281f9 Change the DB query functions to return values of the actual types they are, rather than to return everything as strings -- this will prevent the need for parseInt on COUNT(*) values, etc. and is generally better 
This will temporarily break the display of numeric fields on the interface side
 2006-06-16 21:44:41 +00:00
Dan Stillman
936ef86584 Fix Item.numNotes() and Item.getNotes() to work on items not yet in the DB (return 0 and [], respectively) 2006-06-16 21:33:03 +00:00
Dan Stillman
4904c04e3e Add dateCreated and dateModified columns to itemNotes 
Update itemNotes.dateModified on item update
 2006-06-16 16:09:18 +00:00
Dan Stillman
3f27a12b0f Trigger notify('modify', 'item', noteID) on note modify 
Added method Item.updateDateModified to just change the date -- used on note updates

Delete item notes on item delete
 2006-06-16 15:57:52 +00:00
Dan Stillman
08a570a38a Item.numNotes() 2006-06-16 15:18:01 +00:00
Dan Stillman
47d59f2dc1 Add note contents to search 2006-06-16 08:04:01 +00:00
Dan Stillman
fd85af40f8 Many-to-one item note support in the schema and data layer -- still some issues on (I think) the interface side 2006-06-16 07:32:48 +00:00
Dan Stillman
f635ee7788 Make the retry interval 60 minutes, not 60 seconds 2006-06-15 21:24:04 +00:00
Dan Stillman
8e97675cc9 Added a timer to run repository checks while the browser is up 
Added a separate retry interval so that the extension retries sooner after failures (browser offline, request failure, etc.)

Revision 200 -- w00t i am victorious
 2006-06-15 21:06:24 +00:00
Dan Stillman
2d20fe717d Added Scholar.HTTP.browserIsOffline() and changed doGet() and doPost() to return false if so 2006-06-15 16:52:46 +00:00
Dan Stillman
11a056ecd4 _getDBVersion() caches the version number, so make sure _updateDBVersion() updates it 2006-06-15 16:28:11 +00:00
Dan Stillman
70216ea2c7 - Added automatic scraper update mechanism (more details on Basecamp: http://chnm.grouphub.com/C2687015) 
- Removed localLastUpdated field from scrapers table and renamed centralLastUpdated to lastUpdated; updated scraper queries accordingly

- Added query in scrapers.sql to update version table 'repository' row to prevent immediate downloads of newly installed scrapers

- Get version property from extension manager in Scholar.init() and assign to Scholar.version
 2006-06-15 06:13:02 +00:00
Dan Stillman
8ed37732c1 JS Date() takes months 0-11, for reasons that are far from clear 2006-06-14 16:59:29 +00:00
Dan Stillman
c8a74e96cd Don't be dumb 2006-06-14 15:41:25 +00:00
Dan Stillman
ba16889bb0 Moved the Scholar.Hash constructor out of the main Scholar constructor, where I had put it for some reason -- ignore 2006-06-13 15:14:22 +00:00
Dan Stillman
ca2045a305 Scholar.Date.sqlToDate(string sqldate) -- function to convert SQL-formatted date (e.g. '2006-06-13 11:03:05' or '2006-06-13') into a JS date object 2006-06-13 15:07:08 +00:00
Dan Stillman
39f9d2c3b3 Wrapper functions for XMLHTTPRequest in scholar.js, stolen and adapted slightly from Simon's ingester code (those override the mime type to text/plain and only use responseText, I assume to conform to the Piggy Bank API, so unfortunately we can't use the same ones elsewhere) 
Scholar.HTTP.doGet(url, onStatus, onDone) and Scholar.HTTP.doPost(url, body, onStatus, onDone) -- onStatus and onDone are callbacks to call on non-200 responses and the response body, respectively
 2006-06-13 14:53:38 +00:00
Dan Stillman
d42258b168 Changed schema of scrapers table to use single GUID for scraperID 
Assigned guids to scrapers, replaced INSERT queries with REPLACE queries, and removed table DELETE query at top -- this will allow scrapers to be updated without deleting any others that may exist (e.g. that someone is developing, third-party, etc.)
 2006-06-12 15:43:24 +00:00
Dan Stillman
0b1765b549 Move CreatorTypes constructor in code -- ignore 2006-06-12 13:08:36 +00:00
Dan Stillman
c24c148e0b Cache creator type names to prevent repeated DB lookups in CreatorTypes.getTypeName() 2006-06-12 13:05:30 +00:00
Dan Stillman
5a045f5b50 Cache item type names to prevent repeated DB lookups in ItemTypes.getTypeName() 2006-06-12 12:59:25 +00:00
David Norton
9f7437f5ca [interface] itemsTreeView correctly handles notify(erase) 
[data access] Collection.removeItem() calls trigger() with the right type
 2006-06-09 14:01:07 +00:00
Dan Stillman
9daa0d3303 Added 'notes' field to fields table and sample data (changes are less messy than they look--I just put 'notes' after 'rights' and 'source' and incremented the other fieldIDs) 
scholar.properties addition is just to keep metadata panel from breaking and could probably be removed once interface is hard-coded to not display the notes field there
 2006-06-08 21:30:22 +00:00
Dan Stillman
c3acdf04a6 Don't mark item as changed if setField('') is called on an empty field (i.e. prevent unnecessary save() calls) 2006-06-08 21:08:50 +00:00
Dan Stillman
3715f2c89e - Send a 'modify' to collections in addItem() and removeItem() when collections become or stop being empty 
- Send a 'delete' rather than a 'remove' to itemViews when items are actually deleted (removals from collections still get 'remove' unless it's part of a collection erase)
 2006-06-08 20:41:37 +00:00
Dan Stillman
7c9b1f785c Only send itemView 'remove' notifictions for items if they're actually deleted from the DB, not if they're just removed from a collection 2006-06-08 20:02:26 +00:00
Dan Stillman
01f50b9e4b On Collection.changeParent(), change notify() action to 'move', and always send a three-value array in the order [collectionID, previousParent, newParent] -- 2nd or 3rd param will be null if moving from/to root 2006-06-08 19:58:54 +00:00