source: ndgCommon/trunk/ndg/common/src/dal/ndgRetrieve.py @ 4793

'''
 Given an uri (an instance of ndgObject), retrieve it, with security in place.
 
 @author: B Lawrence?
'''
import cgi,time, logging
import DocumentRetrieve
from ndg.common.src.clients.ws.ndgSearch import ndgSearch
from ndg.common.src.models.ndgObject import ndgObject as no

MOLES_COLLECTION = '/db/ndg_B_metadata'
ATOM_COLLECTION = '/db/atoms'
ATOM_BACKUP_COLLECTION = '/db/atoms_backup'
NUMSIM_COLLECTION = '/db/NumSim'
NDGA0_COLLECTION = '/db/ndg_A_metadata'
DIF_COLLECTION = '/db/DIF'

class NDGRetrieveError(Exception):
    """Exception handling for NDG Retrieve method."""
    def __init__(self, msg):
        logging.error(msg)
        Exception.__init__(self, msg)


def ndgRetrieve(uri, config, output='',remote=0, discovery=1):
    
    ''' Given an uri (an instance of ndgObject), retrieve it, with security in place.
            If output is not '' it should be desired output schema!
            If remote is non zero, then attempt to get the document
        from a remote exist site via a (secured) restful http call
            If discovery is non-zero AND the schema is in 
                [DIF,MDIP,or ISO]
        then use the ndgSearch interface to obtain the document rather
        than direct from an eXist database.'''
    
    def getws(config,uri,remote):
        ''' Get a ws connection to the local exist database '''
        logging.debug("Getting WS connection to eXist DB")
        # The logic here is that
        #    this code is running on server, which supports repositories sr
        #    this particular object needs the repository for uri.repository
        #    which had better be the same as sr, and then we use the
        #    actual exist repository er.
        #    I'm sure this can be cleaned up further :-)
        if remote and discovery:
            raise ValueError, 'Invalid options to ndgRetrieve (remote AND discovery impossible)'
        if discovery:
            logging.debug("Using Discovery service")
            if uri.schema in no.DISCOVERY_SCHEMES:
                ws=ndgSearch()
            else:
                raise NDGRetrieveError('Invalid schema (%s) with discovery \
                    option in ndgRetrieve' %uri.schema)
        else:
            logging.debug("Using Browse service")
            try:
                r=config.get('NDG_B_SERVICE',uri.repository)
                sr=config.get('DEFAULT','repository')
                logging.debug("NDG_B_SERVICE: '%s', default repository: '%s'" %(r,sr))
            except:
                raise NDGRetrieveError("Config file not available or incomplete " + \
                    "- cannot find data for NDG_B_SERVICE, %s" %uri.repository)
            
            if r <> sr:
                if not remote:
                    return 0,'The uri [%s] is not available on [%s]'%(uri,sr)
                else:
                    try:
                        ps=config.get('DEFAULT','proxyServer')
                    except:
                        raise NDGRetrieveError('Config file not available \
                            or has no [proxyServer] setting')
                logging.debug("Retrieving data from remote proxy server, '%s'" %ps)
                ws = DocumentRetrieve.ndgHTTP(r,proxyServer=ps)
            else:
                try:
                    logging.info("Lookup up repository:'%s'" %uri.repository)
                    er=config.get('NDG_EXIST',uri.repository)
                    pwf=config.get('NDG_EXIST','passwordFile')
                except:
                    raise NDGRetrieveError('Config file not available or has \
                        no [NDG_EXIST] setting')

                logging.debug("Retrieving data from repository, '%s'" %er)
                ws=DocumentRetrieve.DocumentRetrieve(er,pwfile=pwf)
        return 1,ws
    
    status,ws=getws(config,uri,remote)
  
    if not status: return status,ws
           
    if uri.schema==no.NDGA0_DOC_TYPE:
        target = NDGA0_COLLECTION
    elif uri.schema == no.NUMSIM_DOC_TYPE:
        target = NUMSIM_COLLECTION
    elif uri.schema == no.ATOM_DOC_TYPE:
        target = ATOM_COLLECTION 
    elif uri.schema == no.ATOM_BACKUP_DOC_TYPE:
        target = ATOM_BACKUP_COLLECTION 
    else:
        target = MOLES_COLLECTION
   
    #do the actual retrieve:
    e=None
    try:
        time1=time.time()
        r=ws.get(uri.repository,uri.schema,uri.localID,targetCollection=target)
        time2=time.time()-time1
        logging.info('Document retrieve [%s] took [%s]'%(uri,time2))
    except Exception,e:
        e=str(e)
        r=e
        logging.error('Document retrieve [%s] failed [%s]'%(uri,e))
    

    # did we get a sensible answer?
    if isinstance(r,int):
        return 0,'<p> There are %s identifiers matching your request! </p>'%r
    if e is not None:
        return 0,'Document retrieve [%s] failed [%s]'%(uri,cgi.escape(e))
    
    # we used to return an xmlHandler instance, but we don't do that any more ...
    return 1,r