3rd party tools for SPIRES
APIs, Mashups, etc.
I want to use some other tools
Many people have written useful tools to help people use SPIRES. When
we know of the tool, we'll provide a link to it, and offer to house the
downloadable code (if appropriate) to preserve it. Remember though, that
these are 3rd party apps, we don't take responsibility for any of them! The following projects
are known (and linked if available):
- Integration with search boxes
- Firefox integration can be done from the browser:
- Go to
the
SPIRES
HEP search page
-
Right click into the search field, select "Add a keyword for this
search..."
-
Give it some name and a keyword (I use spires as the keyword) and save
it to your bookmarks
-
Now you can do a spires search by typing instead of a URL in the location
bar "spires a witten and t jones polynomial".
Ubiquity:
Ubiquity commands
for SPIRES
Command-line tools for perl/python etc
Bibliography services
Things we haven't thought of
If you know of tools like the above, let us know at spires@slac.stanford.edu so
that we can provide links.
I have written/want to write 3rd party tools
SPIRES is happy to have people write useful code to access SPIRES data in
other ways. If possible we'd like to be able to provide a permanaent home
for such code and/or link to a place where it is available. Tools along
the lines of the above are most welcome. If you are writing such tools you may want to contact us at spires@slac.stanford.edu so
that we can provide advice.You might also look at the considerations below
(which are not exhaustive, nor even guaranteed to be up to date.)
Rough guidelines for 3rd party tools
Not exhaustive, nor even guaranteed to be up to date.
- Robot restrictions
- We block robot harvests because it puts
a large load on our servers: If you download large amounts of data rapid
fire and/or indiscriminantly follow links, your IP will be blocked. If
your tool involves making many (more than 10s) searches in a row, think of
a better way, or build in a delay between requests and use our secondary server.
SPIRES is an old system, and can't handle the rapid fire requests that
modern systems might.
- Use our secondary server
- If you must make automated requests
for your service, you might also
try appending &server=sunspi5 to your URLs. This server is used
for robot harvests, and some testing. The data is updated from the real
server nightly, and it occasionally has new (read broken)
functionality before the real server does. However, if you hit this
server hard, our interactive users won't notice, and you'll have less
chance of getting blocked (you still need to follow the rules above, though).
- Understand our URLs
- The only API we have (so far) is our URL syntax.
Playing around on the website is enough to learn the basics of the URL
structure:BaseURL/dbname/format?search&option=value
- BaseURL
- http://www.slac.stanford.edu/spires/find
- dbname
- hep or other database you want to search (abstracts,
hepnames, etc)
- format
- www the name of the format you want to see the results
in. For automated apps, you might try:
- xmlpublic
- rss
- wwwbriefbibtex
- search
- Define the search either as
- rawcmd=url
encoded search from search box
- field=value&field=valuethis will result in
anding the various conditions
- options
- options include
- server
- selects
secondary server (sunspi5)
- format
unneeded (redundant with the
format part of the url)- skip
skips the first N
results
- sequence
- performs an ascii sort on the element listed,
append (d) to the element name to sort descending
-
Examples
- find papers by author books in date 2007, sorted by
eprint number in xml
http://www.slac.stanford.edu/spires/find/hep/xmlpublic?a=brooks&date=2007&sequence=eprint(d)
find papers by babar collaboration in arxiv hep-ex, using the secondary
server, skipping the first 100 results http://www.slac.stanford.edu/spires/find/hep/xmlpublic?rawcmd=find+cn+babar+and+parx+hep-ex&skip=100&server=sunspi5
Use XMLPublic formatThis format is selectable via the url,
but not on the webpages. There is no schema for this format (yet), however, it
is a reasonable selection of elements presented in easy machine readable
XML. No elements will ever be removed from this format, however they may
be added without notice, so any parser should be prepared to find (and
ignore) new things. Note that multiple SPIRES values usually occur mulitply in
the XML, not concatenated in one XML element. Some things that are
usually singly occurring, may occur more than once occasionally. This XML
format isn't perfect, but it is better than most other options. If there
is an element you want to see here, let us know. If there is a
fundamental problem with the xml format, let us know that too.
|
|