In ScrapingStrategy, also check whether found links match a (LATER configurable) compiled regexp and then add to found services the url with an added ending ?WSDL.
Regexp must match : .wso, .asmx
3. LATER add smart (local) spidering :
(first see if there's not an existing lib doing that, or a better way to merge it in ScrapingStrategy)
remember scraped pages (including linked found services)
applications : In ContextStrategy, add a map of (LATER configurable) compiled regexp to replaces, and if browsed URL matches the regexp, apply the replace and scrape it (NB. stay local i.e. no host change) : ^/nuxeo/.* => add webservices/ , ^/alfresco/.* => add api/
LATER engines : try scraping (configurable) /cxf/ , /WS/
by not scraping those links and rather using the top level page url.
Suggest better names in more cases
For now, it only extracts from WSDL the "name". If none, try getting service/@name , portType/@name (or binding/@name) ().
LATER Support interfaces-only or buggy wsdl
See examples at http://www.w3.org/2002/ws/ra/ . Ex. : missing version, web discovery gets back HTML or XSD instead of WSDLs.
LATER with new model, separate interface (signature) from endpoint (and binding = technology).
LATER also get and store .xsd (schemas).
Buggy wsdl :
validate it (with EasyWSDL...) and store this information in the new model
make our tools (SOAPUI, Scaffolder, JWT...) check it but also more robust against buggy wsdl. Ex. for now, SOAPUI explodes on WSDLs from http://www.w3.org/2002/ws/ra/ .
1. add some WS samples to web discovery home :
http://footballpool.dataaccess.eu/data/info.wso => http://footballpool.dataaccess.eu/data/info.wso?WSDL http://webservices.daehosting.com/services/ http://services.unitedplanet.de/blz/BlzService.asmx?WSDL
(and a few others from https://predic8.com/svn/membrane/registry/trunk/servicelist.text )
2. improve scraping :
In ScrapingStrategy, also check whether found links match a (LATER configurable) compiled regexp and then add to found services the url with an added ending ?WSDL. Regexp must match : .wso, .asmx
3. LATER add smart (local) spidering :
Read https://github.com/easysoa/EasySOA/wiki/Service-finder-strategies-reference
support HTTP anchor (#)
by not scraping those links and rather using the top level page url.
Suggest better names in more cases
For now, it only extracts from WSDL the "name". If none, try getting service/@name , portType/@name (or binding/@name) ().
LATER Support interfaces-only or buggy wsdl
See examples at http://www.w3.org/2002/ws/ra/ . Ex. : missing version, web discovery gets back HTML or XSD instead of WSDLs.
LATER with new model, separate interface (signature) from endpoint (and binding = technology). LATER also get and store .xsd (schemas).
Buggy wsdl :