automaton.py 18.2 KB
Newer Older
1
# -*- coding: utf-8 -*-
2
""" harvest_tools.automaton
3 4 5 6 7

"""
import re
import traceback

8

9 10
from base import format_author_fr, MSG_FIX_ORIGIN, MSG_IN_DB, ToolException
from gluon.storage import Storage
11 12 13 14 15 16
from invenio_tools import (CheckAndFix,
                           InvenioStore,
                           Marc12,
                           OAI_URL,
                           REG_OAI,
                           REG_YEAR)
17 18 19 20
from msg import Msg
from msgcollection import MsgCollection
from plugin_dbui import get_create_id, get_id, UNDEF_ID

21

22 23 24
MSG_NO_CAT = 'Select a "category" !!!'
MSG_NO_PROJECT = 'Select a "project" !!!'
MSG_NO_TEAM = 'Select a "team" !!!'
25

26
MSG_NO_OAI = "Reject no OAI identifier"
27
MSG_WELL_FORM_OAI = "Reject OAI is not well formed"
28

29

30
class Automaton(object):
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 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
    """Base class to search and process publications.
        - Decode the parameter of a selector defining user criteria.
        - Search for publications in the store, according to
          user criteria or process and XML string.
        - Load records in the database.

    The parameters of the search as well as the parameters of the harvester
    are defined by the current request.

    """

    def __init__(self,
                 db,
                 id_team,
                 id_project,
                 controller,
                 id_category,
                 year_start=None,
                 year_end=None,
                 dry_run=True,
                 debug=False):
        """
        @note see C{build_harvester_tool} factory function building
              C{PublicationsTools}

        @type db: gluon.dal.DAL
        @param db:

        @type id_team: int
        @param id_team: Identifier of the team in the db

        @type id_project: int
        @param id_project: Identifier of the project in the db

        @type controller: unicode
        @param controller: Type of publication
                           (i.e. 'article', 'proceedings', ...)

        @type id_category: int
        @param id_category: Identifier of the category of publication
                            (i.e. ACL, ACTI, ...)

        @type year_start: int
        @keyword year_start: Start year of search (i.e. '2014')

        @type year_end: int
        @keyword year_end: End year of search (i.e. '2015')

        @type dry_run: boolean
        @keyword dry_run: True if no record is to be written to the db

        @type debug: bool
        @param debug: activate the debug mode

        """
        self.collection_logs = []
        self.db = db
        self.id_team = id_team
        self.id_project = id_project
        self.controller = controller
        self.id_category = id_category
        self.year_start = year_start
        self.year_end = year_end
        self.dry_run = dry_run
        self.dbg = debug
        self.logs = []

        self.check = CheckAndFix()
        self.marc12 = Marc12()

        # check parameters
        # protection team, project and/or category have to be defined
        if not self.id_team:
            raise ToolException(MSG_NO_TEAM)

        if not self.id_project:
            raise ToolException(MSG_NO_PROJECT)

        if not self.id_category:
            raise ToolException(MSG_NO_CAT)

        # Construct harvester Storage needed for the log
        self.harvester = Storage(id_teams=self.id_team,
                                 id_projects=self.id_project,
                                 controller=self.controller,
                                 id_categories=self.id_category)

        # private cache for my_author rescue list
        self.__par = None
        self.__reference = None

122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137
    def _is_in_db(self, rec_id, title):
        """Return C{True} if the record is already in the database.
        The search is based on the origin field.

        A new log entry is created

        @type rec_id: int
        @param rec_id: record identifier

        @type title: str
        @param title: title of the collection

        @rtype: bool

        """
        db = self.db
138
        harvester = self.harvester
139

140 141
        # check
        url = OAI_URL % (harvester.host, rec_id)
142 143 144 145 146 147 148
        db_id = get_id(db.publications, origin=url)

        if db_id is None:
            return False

        publication = db.publications[db_id]

149 150 151 152 153 154 155
        # same category for the publication and the harvester
        # keep the record if it is not the case
        # this is required to transform a preprint into article
        if publication.id_categories != harvester.id_categories:
            return False

        # log
156 157 158 159 160 161 162 163 164
        self.logs.append(Msg(harvester=self.harvester,
                             collection=title,
                             record_id=rec_id,
                             title=publication.title))

        self.logs[-1].idle(MSG_IN_DB, publication.year)

        return True

165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366
    def _search_parameters(self, collection):
        """Build the keywords to steer the URL search in invenio store.
        The main parameter is the collection and the date range defined
        in the selector.

        @type collection: unicode
        @param collection: statement defining the collection in the
        store, I{i.e.} C{"find cn d0 and tc p and not tc c"} or
        C{"LHCb Papers"}. The syntax depends on the invenio store.

        @rtype: dict
        @return: the key are a sub-set of those defined in
        L{invenio_tools.InvenioStore.get_ids}.

        """

        # INSPIREHEP store
        if collection.startswith('find'):

            query = collection

            if self.year_start and not self.year_end:
                query += " and date %s" % self.year_start

            elif not self.year_start and self.year_end:
                query += " and date %s" % self.year_end

            elif self.year_start and self.year_end:
                query += " and date > %s and date < %s " \
                         % (self.year_start - 1, self.year_end + 1)

            dic = dict(p=query,  # query à la spires
                       rg=1000,  # maximum number of records returned
                       sf='year',  # sort by date
                       so='d')  # descending order

        # CERN INVENIO store
        else:

            if self.year_start and not self.year_end:
                rex = self.year_start

            elif not self.year_start and self.year_end:
                rex = self.year_end

            elif self.year_start and self.year_end:
                li = []
                for year in range(self.year_start, self.year_end + 1):
                    li.append(str(year))
                rex = '|'.join(li)

            dic = dict(cc=collection,  # collection
                       f1='year',  # search on year
                       m1='r',  # use regular expression
                       p1=rex,  # regular expression defining year
                       sf='year',  # sort by date
                       so='d')  # descending order
        return dic

    def _my_author_list(self, record):
        """Extract the rescue list for my authors in the database.

        @type record: L{Record}
        @param record:

        @rtype: list
        @return: empty when not defined

        """
        year = record.year()

        # try to recover year when not defined
        if not year:
            # published article, proceeding
            if "773" in record and "y" in record["773"]:
                year = record["773"]["y"]

            # start date of a conference
            elif "111" in record and "x" in record["111"]:
                year = record["111"]["x"]

            # end date of a conference
            elif "111" in record and "z" in record["111"]:
                year = record["111"]["z"]

            # submitted date
            elif "269" in record and "c" in record["269"]:
                year = record["269"]["c"]

            else:
                return []

        #
        # NOTE:
        # keep in mind that the CheckAndfix mechanism is not yet run
        # therefore year can be a list due to erratum, ...
        #
        if isinstance(year, list):
            year.sort()
            year = year[0]

        # the value can have several format 1992, 1992-12-31, ....
        m = REG_YEAR.search(year)
        if m:
            year = m.group(1)

        else:
            return []

        # caching
        t = (year, self.id_project, self.id_team)
        if t == self.__par:
            return self.__reference

        # extract the list from the database
        row = self.db.my_authors(year=year,
                                 id_projects=self.id_project,
                                 id_teams=self.id_team)

        if row:
            self.__reference = row['authors'].split(', ')
        else:
            self.__reference = []

        return self.__reference

    def check_by_fields(self, **kwargs):
        """Check that a record already exist using the fields defined
        in the keyword arguments.

            - Fix the field origin when a match is found.
            - Actions are logged.

        @keyword oai_url: typical value is "http://cds.cern.ch/record/123456"
        @keyword year:

        @note: this method can be customised in inherited class
        to perform dedicated action.

        @rtype: tuple
        @return: the tuple (id, status). The id of the record or None.
        The status is equal to one when the existing record was modified
        zero otherwise

        """
        if self.dbg:
            print "check existing record by fields"

        db = self.db

        # origin can't be used for the search
        oai_url = kwargs["oai_url"]
        del kwargs["oai_url"]

        # look for an existing record
        rec_id = get_id(db.publications, **kwargs)
        if not rec_id:
            return (None, 0)

        # fix origin field
        ok = db.publications[rec_id].origin and \
             db.publications[rec_id].origin == oai_url
        if not ok:
            if not self.dry_run:
                db.publications[rec_id] = dict(origin=oai_url)

            self.logs[-1].modify(MSG_FIX_ORIGIN, kwargs["year"])
            return (rec_id, 1)

        self.logs[-1].idle(MSG_IN_DB, kwargs["year"])
        return (rec_id, 0)

    def check_collaboration(self, value):
        """Check that the collaboration exit in the database, create it if not.

        @type value: str or None
        @param value: the name of the collaboration.

        @rtype: int
        @return: the id of the collaboration, UNDEF_ID when not defined

        """
        if not value:
            return UNDEF_ID

        return get_create_id(self.db.collaborations, collaboration=value)

    def check_publisher(self, value):
        """Check that publisher exit in the database, create it if not.

        @type value: str or None
        @param value: the abbreviation of the publisher name.

        @rtype: int
        @return: the id of the publisher, UNDEF_ID when not defined

        """
        if not value:
            return UNDEF_ID

        return get_create_id(self.db.publishers, abbreviation=value)

367 368 369 370
    def check_record(self, record):
        """Check the content of the record in order to fix non conformities.
        Return False when a non conformities has been found and can not be
        corrected.
371

372
        @note: Some checks depend on the type of publications and have to be
373 374
        implemented in inherited class.

375 376 377 378
        @note: The order of the checks matter. It should be oai,
        temporary record, authors, my authors and then a series of checks
        specific to the publication type.

379 380 381 382
        @type record: L{Record}
        @param record:

        @rtype: bool
383 384
        @return: C{False} when a non conformity is found and can not be
        corrected.
385 386 387

        """
        if self.dbg:
388
            print "check record"
389 390 391 392 393 394 395

        try:
            self.check.temporary_record(record)
            self.check.authors(record)
            self.check.format_authors(record, format_author_fr)
            self.check.collaboration(record)

396
        except Exception as e:
397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
            self.logs[-1].reject(e, record.year())
            return False

        return True

    def load_db(self, record):
        """Load the record in the database.

        @note: This method depend on the type of publications.
        It has to be implemented for each inherited class.

        @type record: L{Record}
        @param record:

        @rtype: int
        @return: one when the record is inserted / updated in the database
        zero otherwise.

        """
        return 0

    def process_xml(self, xml):
        """Decode the xml and load it in the database.

        @raise Exception: the type of exception depends on what happen:
            - L{ToolException} when projet, team or category identifier
              are not defined.
            - C{StoreException} when somethings goes wrong interrogating the
                                store.
            - C{Marc12Exception} when somethings goes wrong decoding the XML
              string return by the store
            - C{CheckException} if the L{Record} is not valid
            - C{Exception} if the python code crash

        @type xml: unicode
        @keyword xml: marc12 xml encoding of the publication record

        """
        if self.dbg:
            print "start processing", self.__class__.__name__
            print "decode request"

        if self.dbg:
            print "get harvest parameters"

        # decode the XML request
        self.collection_logs.append(MsgCollection(found=1))
        self.decode_xml(xml)

    def process_url(self, host, collections):
        """Retrieve the xml from the invenio store and load it in the database

        @raise Exception: depending on what happen, can be StoreException,
        Marc12ZException, ...

        @type host: unicode
        @keyword host: Web host name to query for publication

        @type collections: unicode
        @keyword collections: Request string to send to the host to get the
                              publications

        """
        if self.dbg:
            print "process URL search"

        # extend harvester for logs
        self.harvester.host = host
        self.harvester.collections = collections

        store = InvenioStore(host)

        # list of collections
        collections = re.sub(' *, *', ',', collections).split(',')

        # alias
        controller = self.controller
        project = self.db.projects[self.id_project].project

        # extract the list of publications from the store for each collection
        # the search is perform on a range of creation date
        # if not defined all element are return
        #
        # The method use here minimise the memory usage
        # on the server as well as on the client side

        for collection in collections:

            # log collection information
            # A collection is identified as "Project Controller collection"
            title = "%s / %s / %s" % (project, controller, collection)
            self.collection_logs.append(MsgCollection(title=title))

            # search record in the harvester repository
            kwargs = self._search_parameters(collection)

            try:
                rec_ids = store.get_ids(**kwargs)

            except Exception as error:
                self.collection_logs[-1].url = store.last_search_url()
                self.collection_logs[-1].error = error
                continue

            self.collection_logs[-1].url = store.last_search_url()
            self.collection_logs[-1].found = len(rec_ids)

            if not rec_ids:
                continue

            if self.dbg:
                print '%i records found in %s' % (len(rec_ids), collection)

            for rec_id in rec_ids:

                if self.dbg:
                    print "\nprocessing record", rec_id

                try:
516 517 518
                    if self._is_in_db(rec_id, title):
                        continue

519 520 521
                    xml = store.get_record(rec_id)
                    self.decode_xml(xml)

522
                except Exception as e:
523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556
                    print traceback.format_exc()
                    url = OAI_URL % (host, rec_id)
                    self.logs.append(Msg(harvester=self.harvester,
                                         collection=title,
                                         record_id=rec_id,
                                         title=url))
                    self.logs[-1].reject(e)

    def decode_xml(self, xml):
        """Decode the MARC XML string and load records in the database.

        @type xml: unicode
        @param xml: MARC XML string

        """
        if self.dbg:
            print "process xml record"

        # NOTE: BaseException and inherited class
        # are catched by the previous stage
        records = self.marc12(xml)

        # process individual record
        for record in records:

            if self.dbg:
                print "record decoded"

            # start the log for the record
            self.logs.append(Msg(harvester=self.harvester,
                                 collection=self.collection_logs[-1].title,
                                 record_id=record.id(),
                                 title=record.title()))

557 558 559 560 561 562 563 564 565 566
            # reject record with undefined OAI field
            oai = record.oai()
            if not oai:
                self.logs[-1].reject(MSG_NO_OAI, record.year())

            # reject record is not well form OAI
            match = REG_OAI.match(oai)
            if not match:
                self.logs[-1].reject(MSG_WELL_FORM_OAI, record.year())

567 568 569
            # check that the record is well formed
            # repair non-conformity as far as possible
            if not self.check_record(record):
570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596
                continue

            if self.dbg:
                print "start loading in the database"

            # load record in the database
            self.load_db(record)

            if self.dbg:
                print self.logs[-1].action.upper(), self.logs[-1].txt

    def report(self):
        """Build the processing report.

        @rtype: dict
        @return:
            - C{collection_logs} (list) one L{MsgCollection}) for each
                                        collection
            - C{controller} (str)
            - C{logs} (list) one L{Msg} for each publication
            - C{selector} (Selector)

        """

        return dict(collection_logs=self.collection_logs,
                    controller=self.controller,
                    logs=self.logs)