sporada/nopaque
1
0
Fork 0
mirror of https://gitlab.ub.uni-bielefeld.de/sfb1288inf/nopaque.git synced 2024-11-14 16:55:42 +00:00
Code Issues Projects Releases Wiki Activity

use cqi from pypi

Browse Source
This commit is contained in:
Patrick Jentsch 2020-04-24 15:44:40 +02:00
parent 7fc1084749
commit 98fecacb11
14 changed files with 2 additions and 1690 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
app/corpora/cqi/__init__.py
View File

@ -1,8 +0,0 @@
# flake8: noqa
from .api import APIClient
from .client import CQiClient
from .version import version, version_info
__title__ = 'CQi'
__version__ = version

2
app/corpora/cqi/api/__init__.py
View File

@ -1,2 +0,0 @@
# flake8: noqa
from .client import APIClient

605
app/corpora/cqi/api/client.py
View File

@ -1,605 +0,0 @@
from time import sleep
from . import specification
from ..errors import cl_error_lookup, error_lookup, cqp_error_lookup
import socket
import struct
class APIClient:
"""
A low-level client for the IMS Open Corpus Workbench (CWB) corpus query
interface (CQi) API.
Example:
>>> import cqi
>>> client = cqi.APIClient('127.0.0.1')
>>> client.ctrl_connect('user', 'password')
{'code': 258, 'msg': 'CQI_STATUS_CONNECT_OK'}
>>> client.ctrl_ping()
{'code': 260, 'msg': 'CQI_STATUS_PING_OK'}
>>> client.ctrl_bye()
{'code': 259, 'msg': 'CQI_STATUS_BYE_OK'}
Attributes:
host (str): URL to the CQP server. For example,
``cqpserver.localhost`` or ``127.0.0.1``.
port (int): Port the CQP server listens on. Default: ``4877``
socket (socket.socket): Socket for communicating with a CQP server.
"""
def __init__(self, host, port=4877):
self.host = host
self.port = port
self.socket = socket.socket()
def ctrl_connect(self, username, password):
self.socket.connect((self.host, self.port))
# INPUT: (STRING username, STRING password)
# OUTPUT: CQI_STATUS_CONNECT_OK, CQI_ERROR_CONNECT_REFUSED
self.__send_WORD(specification.CTRL_CONNECT)
self.__send_STRING(username)
self.__send_STRING(password)
return self.__recv_response()
def ctrl_bye(self):
# INPUT: ()
# OUTPUT: CQI_STATUS_BYE_OK
self.__send_WORD(specification.CTRL_BYE)
response = self.__recv_response()
self.socket.close()
return response
def ctrl_user_abort(self):
# INPUT: ()
# OUTPUT:
self.__send_WORD(specification.CTRL_USER_ABORT)
def ctrl_ping(self):
# INPUT: ()
# OUTPUT: CQI_STATUS_PING_OK
self.__send_WORD(specification.CTRL_PING)
return self.__recv_response()
def ctrl_last_general_error(self):
# INPUT: ()
# OUTPUT: CQI_DATA_STRING
# full-text error message for the last general error reported by the
# CQi server
self.__send_WORD(specification.CTRL_LAST_GENERAL_ERROR)
return self.__recv_response()
def ask_feature_cqi_1_0(self):
# INPUT: ()
# OUTPUT: CQI_DATA_BOOL
self.__send_WORD(specification.ASK_FEATURE_CQI_1_0)
return self.__recv_response()
def ask_feature_cl_2_3(self):
# INPUT: ()
# OUTPUT: CQI_DATA_BOOL
self.__send_WORD(specification.ASK_FEATURE_CL_2_3)
return self.__recv_response()
def ask_feature_cqp_2_3(self):
# INPUT: ()
# OUTPUT: CQI_DATA_BOOL
self.__send_WORD(specification.ASK_FEATURE_CL_2_3)
return self.__recv_response()
def corpus_list_coprora(self):
# INPUT: ()
# OUTPUT: CQI_DATA_STRING_LIST
self.__send_WORD(specification.CORPUS_LIST_CORPORA)
return self.__recv_response()
def corpus_charset(self, corpus):
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING
self.__send_WORD(specification.CORPUS_CHARSET)
self.__send_STRING(corpus)
return self.__recv_response()
def corpus_properties(self, corpus):
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING_LIST
self.__send_WORD(specification.CORPUS_PROPERTIES)
self.__send_STRING(corpus)
return self.__recv_response()
def corpus_positional_attributes(self, corpus):
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING_LIST
self.__send_WORD(specification.CORPUS_POSITIONAL_ATTRIBUTES)
self.__send_STRING(corpus)
return self.__recv_response()
def corpus_structural_attributes(self, corpus):
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING_LIST
self.__send_WORD(specification.CORPUS_STRUCTURAL_ATTRIBUTES)
self.__send_STRING(corpus)
return self.__recv_response()
def corpus_structural_attribute_has_values(self, attribute):
# INPUT: (STRING attribute)
# OUTPUT: CQI_DATA_BOOL
self.__send_WORD(specification.CORPUS_STRUCTURAL_ATTRIBUTE_HAS_VALUES)
self.__send_STRING(attribute)
return self.__recv_response()
def corpus_alignment_attributes(self, corpus):
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING_LIST
self.__send_WORD(specification.CORPUS_ALIGNMENT_ATTRIBUTES)
self.__send_STRING(corpus)
return self.__recv_response()
def corpus_full_name(self, corpus):
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING
# the full name of <corpus> as specified in its registry entry
self.__send_WORD(specification.CORPUS_FULL_NAME)
self.__send_STRING(corpus)
return self.__recv_response()
def corpus_info(self, corpus):
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING_LIST
# returns the contents of the .info file of <corpus> as a list of lines
self.__send_WORD(specification.CORPUS_INFO)
self.__send_STRING(corpus)
return self.__recv_response()
def corpus_drop_corpus(self, corpus):
# INPUT: (STRING corpus)
# OUTPUT: CQI_STATUS_OK
# try to unload a corpus and all its attributes from memory
self.__send_WORD(specification.CORPUS_DROP_CORPUS)
self.__send_STRING(corpus)
return self.__recv_response()
def cl_attribute_size(self, attribute):
# INPUT: (STRING attribute)
# OUTPUT: CQI_DATA_INT
# returns the size of <attribute>:
# number of tokens (positional)
# number of regions (structural)
# number of alignments (alignment)
self.__send_WORD(specification.CL_ATTRIBUTE_SIZE)
self.__send_STRING(attribute)
return self.__recv_response()
def cl_lexicon_size(self, attribute):
# INPUT: (STRING attribute)
# OUTPUT: CQI_DATA_INT
# returns the number of entries in the lexicon of a positional
# attribute;
# valid lexicon IDs range from 0 .. (lexicon_size - 1)
self.__send_WORD(specification.CL_LEXICON_SIZE)
self.__send_STRING(attribute)
return self.__recv_response()
def cl_drop_attribute(self, attribute):
# INPUT: (STRING attribute)
# OUTPUT: CQI_STATUS_OK
# unload attribute from memory
self.__send_WORD(specification.CL_DROP_ATTRIBUTE)
self.__send_STRING(attribute)
return self.__recv_response()
"""
" NOTE: simple (scalar) mappings are applied to lists (the returned list
" has exactly the same length as the list passed as an argument)
"""
def cl_str2id(self, attribute, strings):
# INPUT: (STRING attribute, STRING_LIST strings)
# OUTPUT: CQI_DATA_INT_LIST
# returns -1 for every string in <strings> that is not found in the
# lexicon
self.__send_WORD(specification.CL_STR2ID)
self.__send_STRING(attribute)
self.__send_STRING_LIST(strings)
return self.__recv_response()
def cl_id2str(self, attribute, id):
# INPUT: (STRING attribute, INT_LIST id)
# OUTPUT: CQI_DATA_STRING_LIST
# returns "" for every ID in <id> that is out of range
self.__send_WORD(specification.CL_ID2STR)
self.__send_STRING(attribute)
self.__send_INT_LIST(id)
return self.__recv_response()
def cl_id2freq(self, attribute, id):
# INPUT: (STRING attribute, INT_LIST id)
# OUTPUT: CQI_DATA_INT_LIST
# returns 0 for every ID in <id> that is out of range
self.__send_WORD(specification.CL_ID2FREQ)
self.__send_STRING(attribute)
self.__send_INT_LIST(id)
return self.__recv_response()
def cl_cpos2id(self, attribute, cpos):
# INPUT: (STRING attribute, INT_LIST cpos)
# OUTPUT: CQI_DATA_INT_LIST
# returns -1 for every corpus position in <cpos> that is out of range
self.__send_WORD(specification.CL_ID2FREQ)
self.__send_STRING(attribute)
self.__send_INT_LIST(cpos)
return self.__recv_response()
def cl_cpos2str(self, attribute, cpos):
# INPUT: (STRING attribute, INT_LIST cpos)
# OUTPUT: CQI_DATA_STRING_LIST
# returns "" for every corpus position in <cpos> that is out of range
self.__send_WORD(specification.CL_CPOS2STR)
self.__send_STRING(attribute)
self.__send_INT_LIST(cpos)
return self.__recv_response()
def cl_cpos2struc(self, attribute, cpos):
# INPUT: (STRING attribute, INT_LIST cpos)
# OUTPUT: CQI_DATA_INT_LIST
# returns -1 for every corpus position not inside a structure region
self.__send_WORD(specification.CL_CPOS2STRUC)
self.__send_STRING(attribute)
self.__send_INT_LIST(cpos)
return self.__recv_response()
"""
" NOTE: temporary addition for the Euralex2000 tutorial, but should
" probably be included in CQi specs
"""
def cl_cpos2lbound(self, attribute, cpos):
# INPUT: (STRING attribute, INT_LIST cpos)
# OUTPUT: CQI_DATA_INT_LIST
# returns left boundary of s-attribute region enclosing cpos, -1 if not
# in region
self.__send_WORD(specification.CL_CPOS2LBOUND)
self.__send_STRING(attribute)
self.__send_INT_LIST(cpos)
return self.__recv_response()
def cl_cpos2rbound(self, attribute, cpos):
# INPUT: (STRING attribute, INT_LIST cpos)
# OUTPUT: CQI_DATA_INT_LIST
# returns right boundary of s-attribute region enclosing cpos, -1 if
# not in region
self.__send_WORD(specification.CL_CPOS2RBOUND)
self.__send_STRING(attribute)
self.__send_INT_LIST(cpos)
return self.__recv_response()
def cl_cpos2alg(self, attribute, cpos):
# INPUT: (STRING attribute, INT_LIST cpos)
# OUTPUT: CQI_DATA_INT_LIST
# returns -1 for every corpus position not inside an alignment
self.__send_WORD(specification.CL_CPOS2ALG)
self.__send_STRING(attribute)
self.__send_INT_LIST(cpos)
return self.__recv_response()
def cl_struc2str(self, attribute, strucs):
# INPUT: (STRING attribute, INT_LIST strucs)
# OUTPUT: CQI_DATA_STRING_LIST
# returns annotated string values of structure regions in <strucs>; ""
# if out of range
# check CQI_CORPUS_STRUCTURAL_ATTRIBUTE_HAS_VALUES(<attribute>) first
self.__send_WORD(specification.CL_STRUC2STR)
self.__send_STRING(attribute)
self.__send_INT_LIST(strucs)
return self.__recv_response()
"""
" NOTE: the following mappings take a single argument and return multiple
" values, including lists of arbitrary size
"""
def cl_id2cpos(self, attribute, id):
# INPUT: (STRING attribute, INT id)
# OUTPUT: CQI_DATA_INT_LIST
# returns all corpus positions where the given token occurs
self.__send_WORD(specification.CL_ID2CPOS)
self.__send_STRING(attribute)
self.__send_INT(id)
return self.__recv_response()
def cl_idlist2cpos(self, attribute, id_list):
# INPUT: (STRING attribute, INT_LIST id_list)
# OUTPUT: CQI_DATA_INT_LIST
# returns all corpus positions where one of the tokens in <id_list>
# occurs; the returned list is sorted as a whole, not per token id
self.__send_WORD(specification.CL_IDLIST2CPOS)
self.__send_STRING(attribute)
self.__send_INT_LIST(id_list)
return self.__recv_response()
def cl_regex2id(self, attribute, regex):
# INPUT: (STRING attribute, STRING regex)
# OUTPUT: CQI_DATA_INT_LIST
# returns lexicon IDs of all tokens that match <regex>; the returned
# list may be empty (size 0);
self.__send_WORD(specification.CL_REGEX2ID)
self.__send_STRING(attribute)
self.__send_STRING(regex)
return self.__recv_response()
def cl_struc2cpos(self, attribute, struc):
# INPUT: (STRING attribute, INT struc)
# OUTPUT: CQI_DATA_INT_INT
# returns start and end corpus positions of structure region <struc>
self.__send_WORD(specification.CL_STRUC2CPOS)
self.__send_STRING(attribute)
self.__send_INT(struc)
return self.__recv_response()
def cl_alg2cpos(self, attribute, alg):
# INPUT: (STRING attribute, INT alg)
# OUTPUT: CQI_DATA_INT_INT_INT_INT
# returns (src_start, src_end, target_start, target_end)
self.__send_WORD(specification.CL_ALG2CPOS)
self.__send_STRING(attribute)
self.__send_INT(alg)
return self.__recv_response()
def cqp_query(self, mother_corpus, subcorpus_name, query):
# INPUT: (STRING mother_corpus, STRING subcorpus_name, STRING query)
# OUTPUT: CQI_STATUS_OK
# <query> must include the ';' character terminating the query.
self.__send_WORD(specification.CQP_QUERY)
self.__send_STRING(mother_corpus)
self.__send_STRING(subcorpus_name)
self.__send_STRING(query)
return self.__recv_response()
def cqp_list_subcorpora(self, corpus):
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING_LIST
self.__send_WORD(specification.CQP_LIST_SUBCORPORA)
self.__send_STRING(corpus)
return self.__recv_response()
def cqp_subcorpus_size(self, subcorpus):
# INPUT: (STRING subcorpus)
# OUTPUT: CQI_DATA_INT
self.__send_WORD(specification.CQP_SUBCORPUS_SIZE)
self.__send_STRING(subcorpus)
return self.__recv_response()
def cqp_subcorpus_has_field(self, subcorpus, field):
# INPUT: (STRING subcorpus, BYTE field)
# OUTPUT: CQI_DATA_BOOL
self.__send_WORD(specification.CQP_SUBCORPUS_HAS_FIELD)
self.__send_STRING(subcorpus)
self.__send_BYTE(field)
return self.__recv_response()
def cqp_dump_subcorpus(self, subcorpus, field, first, last):
# INPUT: (STRING subcorpus, BYTE field, INT first, INT last)
# OUTPUT: CQI_DATA_INT_LIST
# Dump the values of <field> for match ranges <first> .. <last>
# in <subcorpus>. <field> is one of the CQI_CONST_FIELD_* constants.
self.__send_WORD(specification.CQP_DUMP_SUBCORPUS)
self.__send_STRING(subcorpus)
self.__send_BYTE(field)
self.__send_INT(first)
self.__send_INT(last)
return self.__recv_response()
def cqp_drop_subcorpus(self, subcorpus):
# INPUT: (STRING subcorpus)
# OUTPUT: CQI_STATUS_OK
# delete a subcorpus from memory
self.__send_WORD(specification.CQP_DROP_SUBCORPUS)
self.__send_STRING(subcorpus)
return self.__recv_response()
"""
" NOTE: The following two functions are temporarily included for the
" Euralex 2000 tutorial demo
"""
def cqp_fdist_1(self, subcorpus, cutoff, field, attribute):
""" NOTE: frequency distribution of single tokens """
# INPUT: (STRING subcorpus, INT cutoff, BYTE field, STRING attribute)
# OUTPUT: CQI_DATA_INT_LIST
# returns <n> (id, frequency) pairs flattened into a list of size 2*<n>
# field is one of CQI_CONST_FIELD_MATCH, CQI_CONST_FIELD_TARGET,
# CQI_CONST_FIELD_KEYWORD
# NB: pairs are sorted by frequency desc.
self.__send_WORD(specification.CQP_FDIST_1)
self.__send_STRING(subcorpus)
self.__send_INT(cutoff)
self.__send_BYTE(field)
self.__send_STRING(attribute)
return self.__recv_response()
def cqp_fdist_2(self, subcorpus, cutoff, field1, attribute1, field2,
attribute2):
""" NOTE: frequency distribution of pairs of tokens """
# INPUT: (STRING subcorpus, INT cutoff, BYTE field1, STRING attribute1,
# BYTE field2, STRING attribute2)
# OUTPUT: CQI_DATA_INT_LIST
# returns <n> (id1, id2, frequency) pairs flattened into a list of size
# 3*<n>
# NB: triples are sorted by frequency desc.
self.__send_WORD(specification.CQP_FDIST_2)
self.__send_STRING(subcorpus)
self.__send_INT(cutoff)
self.__send_BYTE(field1)
self.__send_STRING(attribute1)
self.__send_BYTE(field2)
self.__send_STRING(attribute2)
return self.__recv_response()
def __recv_response(self):
byte_data = self.__recv_WORD()
response_type = byte_data >> 8
if response_type == specification.CL_ERROR:
raise cl_error_lookup[byte_data]()
elif response_type == specification.CQP_ERROR:
raise cqp_error_lookup[byte_data]()
elif response_type == specification.DATA:
return self.__recv_DATA(byte_data)
elif response_type == specification.ERROR:
raise error_lookup[byte_data]()
elif response_type == specification.STATUS:
return {'code': byte_data, 'msg': specification.lookup[byte_data]}
else:
raise Exception('Unknown response type: {}'.format(response_type))
def __recv_DATA(self, data_type):
if data_type == specification.DATA_BYTE:
data = self.__recv_DATA_BYTE()
elif data_type == specification.DATA_BOOL:
data = self.__recv_DATA_BOOL()
elif data_type == specification.DATA_INT:
data = self.__recv_DATA_INT()
elif data_type == specification.DATA_STRING:
data = self.__recv_DATA_STRING()
elif data_type == specification.DATA_BYTE_LIST:
data = self.__recv_DATA_BYTE_LIST()
elif data_type == specification.DATA_BOOL_LIST:
data = self.__recv_DATA_BOOL_LIST()
elif data_type == specification.DATA_INT_LIST:
data = self.__recv_DATA_INT_LIST()
elif data_type == specification.DATA_STRING_LIST:
data = self.__recv_DATA_STRING_LIST()
elif data_type == specification.DATA_INT_INT:
data = self.__recv_DATA_INT_INT()
elif data_type == specification.DATA_INT_INT_INT_INT:
data = self.__recv_DATA_INT_INT_INT_INT()
elif data_type == specification.DATA_INT_TABLE:
data = self.__recv_DATA_INT_TABLE()
else:
raise Exception('Unknown data type: {}'.format(data_type))
return data
def __recv_DATA_BYTE(self):
while True:
if (len(self.socket.recv(1, socket.MSG_PEEK)) == 1):
byte_data = self.socket.recv(1)
break
sleep(0.1)
return struct.unpack('!B', byte_data)[0]
def __recv_DATA_BOOL(self):
while True:
if (len(self.socket.recv(1, socket.MSG_PEEK)) == 1):
byte_data = self.socket.recv(1)
break
sleep(0.1)
return struct.unpack('!?', byte_data)[0]
def __recv_DATA_INT(self):
while True:
if (len(self.socket.recv(4, socket.MSG_PEEK)) == 4):
byte_data = self.socket.recv(4)
break
sleep(0.1)
return struct.unpack('!i', byte_data)[0]
def __recv_DATA_STRING(self):
n = self.__recv_WORD()
while True:
if (len(self.socket.recv(n, socket.MSG_PEEK)) == n):
byte_data = self.socket.recv(n)
break
sleep(0.1)
return struct.unpack('!{}s'.format(n), byte_data)[0].decode()
def __recv_DATA_BYTE_LIST(self):
data = []
n = self.__recv_DATA_INT()
while n > 0:
data.append(self.__recv_DATA_BYTE())
n -= 1
return data
def __recv_DATA_BOOL_LIST(self):
data = []
n = self.__recv_DATA_INT()
while n > 0:
data.append(self.__recv_DATA_BOOL())
n -= 1
return data
def __recv_DATA_INT_LIST(self):
data = []
n = self.__recv_DATA_INT()
while n > 0:
data.append(self.__recv_DATA_INT())
n -= 1
return data
def __recv_DATA_STRING_LIST(self):
data = []
n = self.__recv_DATA_INT()
while n > 0:
data.append(self.__recv_DATA_STRING())
n -= 1
return data
def __recv_DATA_INT_INT(self):
return (self.__recv_DATA_INT(), self.__recv_DATA_INT())
def __recv_DATA_INT_INT_INT_INT(self):
return (self.__recv_DATA_INT(),
self.__recv_DATA_INT(),
self.__recv_DATA_INT(),
self.__recv_DATA_INT())
def __recv_DATA_INT_TABLE(self):
rows = self.__recv_DATA_INT()
columns = self.__recv_DATA_INT()
data = []
for i in range(0, rows):
row = []
for j in range(0, columns):
row.append(self.__recv_DATA_INT())
data.append(row)
return data
def __recv_WORD(self):
while True:
if (len(self.socket.recv(2, socket.MSG_PEEK)) == 2):
byte_data = self.socket.recv(2)
break
sleep(0.1)
return struct.unpack('!H', byte_data)[0]
def __send_BYTE(self, byte_data):
data = struct.pack('!B', byte_data)
self.socket.sendall(data)
def __send_BOOL(self, bool_data):
data = struct.pack('!?', bool_data)
self.socket.sendall(data)
def __send_INT(self, int_data):
data = struct.pack('!i', int_data)
self.socket.sendall(data)
def __send_STRING(self, string_data):
encoded_string_data = string_data.encode('utf-8')
n = len(encoded_string_data)
data = struct.pack('!H{}s'.format(n), n, encoded_string_data)
self.socket.sendall(data)
def __send_INT_LIST(self, int_list_data):
n = len(int_list_data)
self.__send_INT(n)
for int_data in int_list_data:
self.__send_INT(int_data)
def __send_STRING_LIST(self, string_list_data):
n = len(string_list_data)
self.__send_INT(n)
for string_data in string_list_data:
self.__send_STRING(string_data)
def __send_WORD(self, word_data):
data = struct.pack('!H', word_data)
self.socket.sendall(data)

404
app/corpora/cqi/api/specification.py
View File

@ -1,404 +0,0 @@
# ########################################################################### #
# IMS CQi specification #
# #
# Version: 0.1a ;o) #
# Author: Stefan Evert (evert@ims.uni-stuttgart.de) #
# Modified by (codestyle): Patrick Jentsch (p.jentsch@uni-bielefeld.de) #
# Modified date: Thurs Oct 10 #
# ########################################################################### #
""" 1. padding """
PAD = 0x00
""" 2. CQi responses """
""" 2.1 CQI_STATUS_* """
STATUS = 0x01
STATUS_OK = 0x0101
STATUS_CONNECT_OK = 0x0102
STATUS_BYE_OK = 0x0103
STATUS_PING_OK = 0x0104
""" 2.2 CQI_ERROR_* """
ERROR = 0x02
ERROR_GENERAL_ERROR = 0x0201
ERROR_CONNECT_REFUSED = 0x0202
ERROR_USER_ABORT = 0x0203
ERROR_SYNTAX_ERROR = 0x0204
# includes corpus/attribute/subcorpus specifier syntax
""" 2.3 CQI_DATA_* """
DATA = 0x03
DATA_BYTE = 0x0301
DATA_BOOL = 0x0302
DATA_INT = 0x0303
DATA_STRING = 0x0304
DATA_BYTE_LIST = 0x0305
DATA_BOOL_LIST = 0x0306
DATA_INT_LIST = 0x0307
DATA_STRING_LIST = 0x0308
DATA_INT_INT = 0x0309
DATA_INT_INT_INT_INT = 0x030A
DATA_INT_TABLE = 0x030B
""" 2.4 CQI_CL_ERROR_* """
"""
" NOTE: some CL error codes are not represented in the CQi specs
" - usually because they're not used in the CL any more
" - CDA_ENOSTRING is not considered an error (returns -1)
" - CDA_EARGS: dynamic attribute calls not yet supported
"""
CL_ERROR = 0x04
CL_ERROR_NO_SUCH_ATTRIBUTE = 0x0401
# returned if CQi server couldn't open attribute
CL_ERROR_WRONG_ATTRIBUTE_TYPE = 0x0402
# CDA_EATTTYPE
CL_ERROR_OUT_OF_RANGE = 0x0403
# CDA_EIDORNG, CDA_EIDXORNG, CDA_EPOSORNG
CL_ERROR_REGEX = 0x0404
# CDA_EPATTERN (not used), CDA_EBADREGEX
CL_ERROR_CORPUS_ACCESS = 0x0405
# CDA_ENODATA
CL_ERROR_OUT_OF_MEMORY = 0x0406
# CDA_ENOMEM
# this means the CQi server has run out of memory;
# try discarding some other corpora and/or subcorpora
CL_ERROR_INTERNAL = 0x0407
# CDA_EOTHER, CDA_ENYI
# this is the classical 'please contact technical support' error
""" 2.5 CQI_CQP_ERROR_* """
CQP_ERROR = 0x05
# CQP error messages yet to be defined
CQP_ERROR_GENERAL = 0x0501
CQP_ERROR_NO_SUCH_CORPUS = 0x0502
CQP_ERROR_INVALID_FIELD = 0x0503
CQP_ERROR_OUT_OF_RANGE = 0x0504
# various cases where a number is out of range
""" 3. CQi commands """
""" 3.1 CQI_CTRL_* """
CTRL = 0x11
CTRL_CONNECT = 0x1101
# INPUT: (STRING username, STRING password)
# OUTPUT: CQI_STATUS_CONNECT_OK, CQI_ERROR_CONNECT_REFUSED
CTRL_BYE = 0x1102
# INPUT: ()
# OUTPUT: CQI_STATUS_BYE_OK
CTRL_USER_ABORT = 0x1103
# INPUT: ()
# OUTPUT:
CTRL_PING = 0x1104
# INPUT: ()
# OUTPUT: CQI_STATUS_PING_OK
CTRL_LAST_GENERAL_ERROR = 0x1105
# INPUT: ()
# OUTPUT: CQI_DATA_STRING
# full-text error message for the last general error reported by the CQi server
""" 3.2 CQI_ASK_FEATURE_* """
ASK_FEATURE = 0x12
ASK_FEATURE_CQI_1_0 = 0x1201
# INPUT: ()
# OUTPUT: CQI_DATA_BOOL
ASK_FEATURE_CL_2_3 = 0x1202
# INPUT: ()
# OUTPUT: CQI_DATA_BOOL
ASK_FEATURE_CQP_2_3 = 0x1203
# INPUT: ()
# OUTPUT: CQI_DATA_BOOL
""" 3.3 CQI_CORPUS_* """
CORPUS = 0x13
CORPUS_LIST_CORPORA = 0x1301
# INPUT: ()
# OUTPUT: CQI_DATA_STRING_LIST
CORPUS_CHARSET = 0x1303
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING
CORPUS_PROPERTIES = 0x1304
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING_LIST
CORPUS_POSITIONAL_ATTRIBUTES = 0x1305
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING_LIST
CORPUS_STRUCTURAL_ATTRIBUTES = 0x1306
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING_LIST
CORPUS_STRUCTURAL_ATTRIBUTE_HAS_VALUES = 0x1307
# INPUT: (STRING attribute)
# OUTPUT: CQI_DATA_BOOL
CORPUS_ALIGNMENT_ATTRIBUTES = 0x1308
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING_LIST
CORPUS_FULL_NAME = 0x1309
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING
# the full name of <corpus> as specified in its registry entry
CORPUS_INFO = 0x130A
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING_LIST
# returns the contents of the .info file of <corpus> as a list of lines
CORPUS_DROP_CORPUS = 0x130B
# INPUT: (STRING corpus)
# OUTPUT: CQI_STATUS_OK
# try to unload a corpus and all its attributes from memory
""" 3.4 CQI_CL_* """
CL = 0x14
# low-level corpus access (CL functions)
CL_ATTRIBUTE_SIZE = 0x1401
# INPUT: (STRING attribute)
# OUTPUT: CQI_DATA_INT
# returns the size of <attribute>:
# - number of tokens (positional)
# - number of regions (structural)
# - number of alignments (alignment)
CL_LEXICON_SIZE = 0x1402
# INPUT: (STRING attribute)
# OUTPUT: CQI_DATA_INT
# returns the number of entries in the lexicon of a positional attribute;
# valid lexicon IDs range from 0 .. (lexicon_size - 1)
CL_DROP_ATTRIBUTE = 0x1403
# INPUT: (STRING attribute)
# OUTPUT: CQI_STATUS_OK
# unload attribute from memory
"""
" NOTE: simple (scalar) mappings are applied to lists (the returned list has
" exactly the same length as the list passed as an argument)
"""
CL_STR2ID = 0x1404
# INPUT: (STRING attribute, STRING_LIST strings)
# OUTPUT: CQI_DATA_INT_LIST
# returns -1 for every string in <strings> that is not found in the lexicon
CL_ID2STR = 0x1405
# INPUT: (STRING attribute, INT_LIST id)
# OUTPUT: CQI_DATA_STRING_LIST
# returns "" for every ID in <id> that is out of range
CL_ID2FREQ = 0x1406
# INPUT: (STRING attribute, INT_LIST id)
# OUTPUT: CQI_DATA_INT_LIST
# returns 0 for every ID in <id> that is out of range
CL_CPOS2ID = 0x1407
# INPUT: (STRING attribute, INT_LIST cpos)
# OUTPUT: CQI_DATA_INT_LIST
# returns -1 for every corpus position in <cpos> that is out of range
CL_CPOS2STR = 0x1408
# INPUT: (STRING attribute, INT_LIST cpos)
# OUTPUT: CQI_DATA_STRING_LIST
# returns "" for every corpus position in <cpos> that is out of range
CL_CPOS2STRUC = 0x1409
# INPUT: (STRING attribute, INT_LIST cpos)
# OUTPUT: CQI_DATA_INT_LIST
# returns -1 for every corpus position not inside a structure region
"""
" NOTE: temporary addition for the Euralex2000 tutorial, but should probably be
" included in CQi specs
"""
CL_CPOS2LBOUND = 0x1420
# INPUT: (STRING attribute, INT_LIST cpos)
# OUTPUT: CQI_DATA_INT_LIST
# returns left boundary of s-attribute region enclosing cpos, -1 if not in
# region
CL_CPOS2RBOUND = 0x1421
# INPUT: (STRING attribute, INT_LIST cpos)
# OUTPUT: CQI_DATA_INT_LIST
# returns right boundary of s-attribute region enclosing cpos, -1 if not in
# region
CL_CPOS2ALG = 0x140A
# INPUT: (STRING attribute, INT_LIST cpos)
# OUTPUT: CQI_DATA_INT_LIST
# returns -1 for every corpus position not inside an alignment
CL_STRUC2STR = 0x140B
# INPUT: (STRING attribute, INT_LIST strucs)
# OUTPUT: CQI_DATA_STRING_LIST
# returns annotated string values of structure regions in <strucs>; "" if out
# of range
# check CQI_CORPUS_STRUCTURAL_ATTRIBUTE_HAS_VALUES(<attribute>) first
"""
" NOTE: the following mappings take a single argument and return multiple
" values, including lists of arbitrary size
"""
CL_ID2CPOS = 0x140C
# INPUT: (STRING attribute, INT id)
# OUTPUT: CQI_DATA_INT_LIST
# returns all corpus positions where the given token occurs
CL_IDLIST2CPOS = 0x140D
# INPUT: (STRING attribute, INT_LIST id_list)
# OUTPUT: CQI_DATA_INT_LIST
# returns all corpus positions where one of the tokens in <id_list>
# occurs; the returned list is sorted as a whole, not per token id
CL_REGEX2ID = 0x140E
# INPUT: (STRING attribute, STRING regex)
# OUTPUT: CQI_DATA_INT_LIST
# returns lexicon IDs of all tokens that match <regex>; the returned
# list may be empty (size 0);
CL_STRUC2CPOS = 0x140F
# INPUT: (STRING attribute, INT struc)
# OUTPUT: CQI_DATA_INT_INT
# returns start and end corpus positions of structure region <struc>
CL_ALG2CPOS = 0x1410
# INPUT: (STRING attribute, INT alg)
# OUTPUT: CQI_DATA_INT_INT_INT_INT
# returns (src_start, src_end, target_start, target_end)
""" 3.5 CQI_CQP_* """
CQP = 0x15
CQP_QUERY = 0x1501
# INPUT: (STRING mother_corpus, STRING subcorpus_name, STRING query)
# OUTPUT: CQI_STATUS_OK
# <query> must include the ';' character terminating the query.
CQP_LIST_SUBCORPORA = 0x1502
# INPUT: (STRING corpus)
# OUTPUT: CQI_DATA_STRING_LIST
CQP_SUBCORPUS_SIZE = 0x1503
# INPUT: (STRING subcorpus)
# OUTPUT: CQI_DATA_INT
CQP_SUBCORPUS_HAS_FIELD = 0x1504
# INPUT: (STRING subcorpus, BYTE field)
# OUTPUT: CQI_DATA_BOOL
CQP_DUMP_SUBCORPUS = 0x1505
# INPUT: (STRING subcorpus, BYTE field, INT first, INT last)
# OUTPUT: CQI_DATA_INT_LIST
# Dump the values of <field> for match ranges <first> .. <last> in <subcorpus>.
# <field> is one of the CQI_CONST_FIELD_* constants.
CQP_DROP_SUBCORPUS = 0x1509
# INPUT: (STRING subcorpus)
# OUTPUT: CQI_STATUS_OK
# delete a subcorpus from memory
"""
" NOTE: The following two functions are temporarily included for the Euralex
" 2000 tutorial demo
"""
""" NOTE: frequency distribution of single tokens """
CQP_FDIST_1 = 0x1510
# INPUT: (STRING subcorpus, INT cutoff, BYTE field, STRING attribute)
# OUTPUT: CQI_DATA_INT_LIST
# returns <n> (id, frequency) pairs flattened into a list of size 2*<n>
# field is one of CQI_CONST_FIELD_MATCH, CQI_CONST_FIELD_TARGET,
# CQI_CONST_FIELD_KEYWORD
# NB: pairs are sorted by frequency desc.
""" NOTE: frequency distribution of pairs of tokens """
CQP_FDIST_2 = 0x1511
# INPUT: (STRING subcorpus, INT cutoff, BYTE field1, STRING attribute1,
# BYTE field2, STRING attribute2)
# OUTPUT: CQI_DATA_INT_LIST
# returns <n> (id1, id2, frequency) pairs flattened into a list of size 3*<n>
# NB: triples are sorted by frequency desc.
""" 4. Constant Definitions """
CONST_FALSE = 0x00
CONST_NO = 0x00
CONST_TRUE = 0x01
CONST_YES = 0x01
"""
" NOTE: The following constants specify which field will be returned by
" CQI_CQP_DUMP_SUBCORPUS and some other subcorpus commands.
"""
CONST_FIELD_MATCH = 0x10
CONST_FIELD_MATCHEND = 0x11
"""
" NOTE: The constants specifiying target0 .. target9 are guaranteed to have the
" numerical values 0 .. 9, so clients do not need to look up the constant
" values if they're handling arbitrary targets.
"""
CONST_FIELD_TARGET_0 = 0x00
CONST_FIELD_TARGET_1 = 0x01
CONST_FIELD_TARGET_2 = 0x02
CONST_FIELD_TARGET_3 = 0x03
CONST_FIELD_TARGET_4 = 0x04
CONST_FIELD_TARGET_5 = 0x05
CONST_FIELD_TARGET_6 = 0x06
CONST_FIELD_TARGET_7 = 0x07
CONST_FIELD_TARGET_8 = 0x08
CONST_FIELD_TARGET_9 = 0x09
"""
" NOTE: The following constants are provided for backward compatibility with
" traditional CQP field names & while the generalised target concept
" isn't yet implemented in the CQPserver.
"""
CONST_FIELD_TARGET = 0x00
CONST_FIELD_KEYWORD = 0x09
""" NOTE: CQi version is CQI_MAJOR_VERSION.CQI_MINOR_VERSION """
MAJOR_VERSION = 0x00
MINOR_VERSION = 0x01
""" 5. CQi lookup dictionary. """
lookup = {
257: 'CQI_STATUS_OK',
258: 'CQI_STATUS_CONNECT_OK',
259: 'CQI_STATUS_BYE_OK',
260: 'CQI_STATUS_PING_OK',
513: 'CQI_ERROR_GENERAL_ERROR',
514: 'CQI_ERROR_CONNECT_REFUSED',
515: 'CQI_ERROR_USER_ABORT',
516: 'CQI_ERROR_SYNTAX_ERROR',
769: 'CQI_DATA_BYTE',
770: 'CQI_DATA_BOOL',
771: 'CQI_DATA_INT',
772: 'CQI_DATA_STRING',
773: 'CQI_DATA_BYTE_LIST',
774: 'CQI_DATA_BOOL_LIST',
775: 'CQI_DATA_INT_LIST',
776: 'CQI_DATA_STRING_LIST',
777: 'CQI_DATA_INT_INT',
778: 'CQI_DATA_INT_INT_INT_INT',
779: 'CQI_DATA_INT_TABLE',
1025: 'CQI_CL_ERROR_NO_SUCH_ATTRIBUTE',
1026: 'CQI_CL_ERROR_WRONG_ATTRIBUTE_TYPE',
1027: 'CQI_CL_ERROR_OUT_OF_RANGE',
1028: 'CQI_CL_ERROR_REGEX',
1029: 'CQI_CL_ERROR_CORPUS_ACCESS',
1030: 'CQI_CL_ERROR_OUT_OF_MEMORY',
1031: 'CQI_CL_ERROR_INTERNAL',
1281: 'CQI_CQP_ERROR_GENERAL',
1282: 'CQI_CQP_ERROR_NO_SUCH_CORPUS',
1283: 'CQI_CQP_ERROR_INVALID_FIELD',
1284: 'CQI_CQP_ERROR_OUT_OF_RANGE',
4353: 'CQI_CTRL_CONNECT',
4354: 'CQI_CTRL_BYE',
4355: 'CQI_CTRL_USER_ABORT',
4356: 'CQI_CTRL_PING',
4357: 'CQI_CTRL_LAST_GENERAL_ERROR',
4609: 'CQI_ASK_FEATURE_CQI_1_0',
4610: 'CQI_ASK_FEATURE_CL_2_3',
4611: 'CQI_ASK_FEATURE_CQP_2_3',
4865: 'CQI_CORPUS_LIST_CORPORA',
4867: 'CQI_CORPUS_CHARSET',
4868: 'CQI_CORPUS_PROPERTIES',
4869: 'CQI_CORPUS_POSITIONAL_ATTRIBUTES',
4870: 'CQI_CORPUS_STRUCTURAL_ATTRIBUTES',
4871: 'CQI_CORPUS_STRUCTURAL_ATTRIBUTE_HAS_VALUES',
4872: 'CQI_CORPUS_ALIGNMENT_ATTRIBUTES',
4873: 'CQI_CORPUS_FULL_NAME',
4874: 'CQI_CORPUS_INFO',
4875: 'CQI_CORPUS_DROP_CORPUS',
5121: 'CQI_CL_ATTRIBUTE_SIZE',
5122: 'CQI_CL_LEXICON_SIZE',
5123: 'CQI_CL_DROP_ATTRIBUTE',
5124: 'CQI_CL_STR2ID',
5125: 'CQI_CL_ID2STR',
5126: 'CQI_CL_ID2FREQ',
5127: 'CQI_CL_CPOS2ID',
5128: 'CQI_CL_CPOS2STR',
5129: 'CQI_CL_CPOS2STRUC',
5130: 'CQI_CL_CPOS2ALG',
5131: 'CQI_CL_STRUC2STR',
5132: 'CQI_CL_ID2CPOS',
5133: 'CQI_CL_IDLIST2CPOS',
5134: 'CQI_CL_REGEX2ID',
5135: 'CQI_CL_STRUC2CPOS',
5136: 'CQI_CL_ALG2CPOS',
5152: 'CQI_CL_CPOS2LBOUND',
5153: 'CQI_CL_CPOS2RBOUND',
5377: 'CQI_CQP_QUERY',
5378: 'CQI_CQP_LIST_SUBCORPORA',
5379: 'CQI_CQP_SUBCORPUS_SIZE',
5380: 'CQI_CQP_SUBCORPUS_HAS_FIELD',
5381: 'CQI_CQP_DUMP_SUBCORPUS',
5385: 'CQI_CQP_DROP_SUBCORPUS',
5392: 'CQI_CQP_FDIST_1',
5393: 'CQI_CQP_FDIST_2'
}

46
app/corpora/cqi/client.py
View File

@ -1,46 +0,0 @@
from .api import APIClient
from .models.corpora import CorpusCollection
class CQiClient:
"""
A client for communicating with a CQi server.
Example:
>>> import cqi
>>> client = cqi.CQiClient('127.0.0.1')
>>> client.connect()
{'code': 258, 'msg': 'CQI_STATUS_CONNECT_OK'}
>>> client.ping()
{'code': 260, 'msg': 'CQI_STATUS_PING_OK'}
>>> client.disconnect()
{'code': 259, 'msg': 'CQI_STATUS_BYE_OK'}
Attributes:
api (APIClient): An API client pointing to the specified CQP server.
"""
def __init__(self, host, port=4877):
"""
CQiClient constructor
Args:
host (str): URL to the CQP server. For example,
``cqpserver.localhost`` or ``127.0.0.1``.
port (int): Port the CQP server listens on. Default: ``4877``
"""
self.api = APIClient(host, port=port)
def connect(self, username='anonymous', password=''):
status = self.api.ctrl_connect(username, password)
return status
def disconnect(self):
return self.api.ctrl_bye()
def ping(self):
return self.api.ctrl_ping()
@property
def corpora(self):
return CorpusCollection(client=self)

180
app/corpora/cqi/errors.py
View File

@ -1,180 +0,0 @@
from .api import specification
class CQiException(Exception):
"""
A base class from which all other exceptions inherit.
If you want to catch all errors that the CQi package might raise,
catch this base exception.
"""
def __init__(self, *args, **kwargs):
super(CQiException, self).__init__(*args, **kwargs)
self.code = None
self.name = None
self.description = None
class Error(CQiException):
def __init__(self, *args, **kwargs):
super(Error, self).__init__(*args, **kwargs)
self.code = specification.ERROR
class ErrorGeneralError(Error):
def __init__(self, *args, **kwargs):
super(ErrorGeneralError, self).__init__(*args, **kwargs)
self.code = specification.ERROR_GENERAL_ERROR
self.name = specification.lookup[self.code]
class ErrorConnectRefused(Error):
def __init__(self, *args, **kwargs):
super(ErrorConnectRefused, self).__init__(*args, **kwargs)
self.code = specification.ERROR_CONNECT_REFUSED
self.name = specification.lookup[self.code]
class ErrorUserAbort(Error):
def __init__(self, *args, **kwargs):
super(ErrorUserAbort, self).__init__(*args, **kwargs)
self.code = specification.ERROR_USER_ABORT
self.name = specification.lookup[self.code]
class ErrorSyntaxError(Error):
def __init__(self, *args, **kwargs):
super(ErrorSyntaxError, self).__init__(*args, **kwargs)
self.code = specification.ERROR_SYNTAX_ERROR
self.name = specification.lookup[self.code]
class CLError(CQiException):
def __init__(self, *args, **kwargs):
super(CLError, self).__init__(*args, **kwargs)
self.code = specification.CL_ERROR
class CLErrorNoSuchAttribute(CLError):
def __init__(self, *args, **kwargs):
super(CLErrorNoSuchAttribute, self).__init__(*args, **kwargs)
self.code = specification.CL_ERROR_NO_SUCH_ATTRIBUTE
self.name = specification.lookup[self.code]
self.description = "CQi server couldn't open attribute"
class CLErrorWrongAttributeType(CLError):
# CDA_EATTTYPE
def __init__(self, *args, **kwargs):
super(CLErrorWrongAttributeType, self).__init__(*args, **kwargs)
self.code = specification.CL_ERROR_WRONG_ATTRIBUTE_TYPE
self.name = specification.lookup[self.code]
class CLErrorOutOfRange(CLError):
# CDA_EIDORNG, CDA_EIDXORNG, CDA_EPOSORNG
def __init__(self, *args, **kwargs):
super(CLErrorOutOfRange, self).__init__(*args, **kwargs)
self.code = specification.CL_ERROR_OUT_OF_RANGE
self.name = specification.lookup[self.code]
class CLErrorRegex(CLError):
# CDA_EPATTERN (not used), CDA_EBADREGEX
def __init__(self, *args, **kwargs):
super(CLErrorRegex, self).__init__(*args, **kwargs)
self.code = specification.CL_ERROR_REGEX
self.name = specification.lookup[self.code]
class CLErrorCorpusAccess(CLError):
# CDA_ENODATA
def __init__(self, *args, **kwargs):
super(CLErrorCorpusAccess, self).__init__(*args, **kwargs)
self.code = specification.CL_ERROR_CORPUS_ACCESS
self.name = specification.lookup[self.code]
class CLErrorOutOfMemory(CLError):
# CDA_ENOMEM
def __init__(self, *args, **kwargs):
super(CLErrorOutOfMemory, self).__init__(*args, **kwargs)
self.code = specification.CL_ERROR_OUT_OF_MEMORY
self.name = specification.lookup[self.code]
self.description = ('CQi server has run out of memory; try discarding '
'some other corpora and/or subcorpora')
class CLErrorInternal(CLError):
# CDA_EOTHER, CDA_ENYI
def __init__(self, *args, **kwargs):
super(CLErrorInternal, self).__init__(*args, **kwargs)
self.code = specification.CL_ERROR_INTERNAL
self.name = specification.lookup[self.code]
self.description = "Classical 'please contact technical support' error"
class CQPError(CQiException):
# CQP error messages yet to be defined
def __init__(self, *args, **kwargs):
super(CQPError, self).__init__(*args, **kwargs)
self.code = specification.CQP_ERROR
class CQPErrorGeneral(CQPError):
def __init__(self, *args, **kwargs):
super(CQPErrorGeneral, self).__init__(*args, **kwargs)
self.code = specification.CQP_ERROR_GENERAL
self.name = specification.lookup[self.code]
class CQPErrorNoSuchCorpus(CQPError):
def __init__(self, *args, **kwargs):
super(CQPErrorNoSuchCorpus, self).__init__(*args, **kwargs)
self.code = specification.CQP_ERROR_NO_SUCH_CORPUS
self.name = specification.lookup[self.code]
class CQPErrorInvalidField(CQPError):
def __init__(self, *args, **kwargs):
super(CQPErrorInvalidField, self).__init__(*args, **kwargs)
self.code = specification.CQP_ERROR_INVALID_FIELD
self.name = specification.lookup[self.code]
class CQPErrorOutOfRange(CQPError):
def __init__(self, *args, **kwargs):
super(CQPErrorOutOfRange, self).__init__(*args, **kwargs)
self.code = specification.CQP_ERROR_OUT_OF_RANGE
self.name = specification.lookup[self.code]
self.description = 'A number is out of range'
error_lookup = {
specification.ERROR: Error,
specification.ERROR_GENERAL_ERROR: ErrorGeneralError,
specification.ERROR_CONNECT_REFUSED: ErrorConnectRefused,
specification.ERROR_USER_ABORT: ErrorUserAbort,
specification.ERROR_SYNTAX_ERROR: ErrorSyntaxError
}
cl_error_lookup = {
specification.CL_ERROR: CLError,
specification.CL_ERROR_NO_SUCH_ATTRIBUTE: CLErrorNoSuchAttribute,
specification.CL_ERROR_WRONG_ATTRIBUTE_TYPE: CLErrorWrongAttributeType,
specification.CL_ERROR_OUT_OF_RANGE: CLErrorOutOfRange,
specification.CL_ERROR_REGEX: CLErrorRegex,
specification.CL_ERROR_CORPUS_ACCESS: CLErrorCorpusAccess,
specification.CL_ERROR_OUT_OF_MEMORY: CLErrorOutOfMemory,
specification.CL_ERROR_INTERNAL: CLErrorInternal
}
cqp_error_lookup = {
specification.CQP_ERROR: CQPError,
specification.CQP_ERROR_GENERAL: CQPErrorGeneral,
specification.CQP_ERROR_NO_SUCH_CORPUS: CQPErrorNoSuchCorpus,
specification.CQP_ERROR_INVALID_FIELD: CQPErrorInvalidField,
specification.CQP_ERROR_OUT_OF_RANGE: CQPErrorOutOfRange
}

0
app/corpora/cqi/models/__init__.py
View File

162
app/corpora/cqi/models/attributes.py
View File

@ -1,162 +0,0 @@
from .ressource import Collection, Model
class Attribute(Model):
"""
This is a class representing an attribute. Attributes denote the general
category of information. A specific occurence is identified by an Id.
"""
id_attribute = 'api_name'
@staticmethod
def _attrs(client, corpus, name):
api_name = '{}.{}'.format(corpus.attrs['api_name'], name)
return {'api_name': api_name,
'name': name,
'size': client.api.cl_attribute_size(api_name)}
def drop(self):
return self.client.api.cl_drop_attribute(self.attrs['api_name'])
class AttributeCollection(Collection):
model = Attribute
def __init__(self, client=None, corpus=None):
super(AttributeCollection, self).__init__(client=client)
self.corpus = corpus
def get(self, attribute_name):
return self.prepare_model(self.model._attrs(self.client, self.corpus,
attribute_name))
def list(self):
raise NotImplementedError
class AlignmentAttribute(Attribute):
def cpos_by_ids(self, id_list):
return self.client.api.cl_alg2cpos(self.attrs['api_name'], id_list)
def ids_by_cpos(self, cpos_list):
return self.client.api.cl_cpos2alg(self.attrs['api_name'], cpos_list)
class AlignmentAttributeCollection(AttributeCollection):
model = AlignmentAttribute
def list(self):
return [self.prepare_model(self.model._attrs(self.client, self.corpus, attr)) # noqa
for attr in self.client.api.corpus_alignment_attributes(self.corpus.attrs['api_name'])] # noqa
class PositionalAttribute(Attribute):
@staticmethod
def _attrs(client, corpus, name):
attrs = super(PositionalAttribute, PositionalAttribute)._attrs(client, corpus, name) # noqa
attrs['lexicon_size'] = client.api.cl_lexicon_size(attrs['api_name'])
return attrs
def cpos_by_id(self, id):
return self.client.api.cl_id2cpos(self.attrs['api_name'], id)
def cpos_by_ids(self, id_list):
return self.client.api.cl_idlist2cpos(self.attrs['api_name'], id_list)
def freqs_by_ids(self, id_list):
return self.client.api.cl_id2freq(self.attrs['api_name'], id_list)
def ids_by_cpos(self, cpos_list):
return self.client.api.cl_cpos2id(self.attrs['api_name'], cpos_list)
def ids_by_regex(self, regex):
return self.client.api.cl_regex2id(self.attrs['api_name'], regex)
def ids_by_values(self, value_list):
return self.client.api.cl_str2id(self.attrs['api_name'], value_list)
def values_by_cpos(self, cpos_list):
return self.client.api.cl_cpos2str(self.attrs['api_name'], cpos_list)
def values_by_ids(self, id_list):
return self.client.api.cl_id2str(self.attrs['api_name'], id_list)
class PositionalAttributeCollection(AttributeCollection):
model = PositionalAttribute
def list(self):
return [self.prepare_model(self.model._attrs(self.client, self.corpus, attr)) # noqa
for attr in self.client.api.corpus_positional_attributes(self.corpus.attrs['api_name'])] # noqa
class StructuralAttribute(Attribute):
@staticmethod
def _attrs(client, corpus, name):
attrs = super(StructuralAttribute, StructuralAttribute)._attrs(client, corpus, name) # noqa
attrs['has_values'] = client.api.corpus_structural_attribute_has_values(attrs['api_name']) # noqa
return attrs
def cpos_by_id(self, id):
return self.client.api.cl_struc2cpos(self.attrs['api_name'], id)
def ids_by_cpos(self, cpos_list):
return self.client.api.cl_cpos2struc(self.attrs['api_name'], cpos_list)
def lbound_by_cpos(self, cpos_list):
return self.client.api.cl_cpos2lbound(self.attrs['api_name'],
cpos_list)
def rbound_by_cpos(self, cpos_list):
return self.client.api.cl_cpos2rbound(self.attrs['api_name'],
cpos_list)
def values_by_ids(self, id_list):
return self.client.api.cl_struc2str(self.attrs['api_name'], id_list)
def export(self, first_cpos, last_cpos, context=0, expand_lists=False):
first_id, last_id = self.ids_by_cpos([first_cpos, last_cpos])
c = (first_cpos, last_cpos)
lc = rc = None
if context == 0:
cpos_list = list(range(first_cpos, (last_cpos + 1)))
else:
lc_lbound = self.cpos_by_id(max(0, (first_id - context)))[0]
if lc_lbound != first_cpos:
lc_rbound = max(0, (first_cpos - 1))
lc = (lc_lbound, lc_rbound)
cpos_list_lbound = lc_lbound
else:
cpos_list_lbound = first_cpos
rc_rbound = \
self.cpos_by_id(min((last_id + context),
(self.attrs['size'] - 1)))[1]
if rc_rbound != last_cpos:
rc_lbound = min((last_cpos + 1),
(self.collection.corpus.attrs['size'] - 1))
rc = (rc_lbound, rc_rbound)
cpos_list_rbound = rc_rbound
else:
cpos_list_rbound = last_cpos
cpos_list = list(range(cpos_list_lbound, (cpos_list_rbound + 1)))
if expand_lists:
match = {'lc': list(range(lc[0], (lc[1] + 1))),
'c': list(range(c[0], (c[1] + 1))),
'rc': list(range(rc[0], (rc[1] + 1)))}
else:
match = {'lc': lc, 'c': c, 'rc': rc}
lookups = self.collection.corpus.lookups_by_cpos(cpos_list)
return {'match': match, **lookups}
class StructuralAttributeCollection(AttributeCollection):
model = StructuralAttribute
def list(self, filters={}):
attrs = [self.prepare_model(self.model._attrs(self.client, self.corpus, attr)) # noqa
for attr in self.client.api.corpus_structural_attributes(self.corpus.attrs['api_name'])] # noqa
for k, v in filters.items():
if k == 'part_of':
attrs = list(filter(lambda x: x.attrs['name'].startswith(v.attrs['name'] + '_'), attrs)) # noqa
return attrs

96
app/corpora/cqi/models/corpora.py
View File

@ -1,96 +0,0 @@
from .attributes import (AlignmentAttributeCollection,
PositionalAttributeCollection,
StructuralAttributeCollection)
from .ressource import Collection, Model
from .subcorpora import SubcorpusCollection
class Corpus(Model):
id_attribute = 'api_name'
@staticmethod
def _attrs(client, name):
api_name = name
return {'api_name': api_name,
'name': name,
'size': client.api.cl_attribute_size(
'{}.word'.format(api_name)),
# 'info': client.api.corpus_info(name),
'charset': client.api.corpus_charset(api_name),
# 'full_name' = client.api.corpus_full_name(name),
'properties': client.api.corpus_properties(api_name)}
def lookups_by_cpos(self, cpos_list):
cpos_list = list(set(cpos_list))
lookups = {}
if cpos_list:
lookups['cpos_lookup'] = {}
for cpos in cpos_list:
lookups['cpos_lookup'][cpos] = {}
for attr in self.positional_attributes.list():
cpos_attr_values = attr.values_by_cpos(cpos_list)
for i, cpos in enumerate(cpos_list):
lookups['cpos_lookup'][cpos][attr.attrs['name']] = \
cpos_attr_values[i]
for attr in self.structural_attributes.list():
if attr.attrs['has_values']:
continue
cpos_attr_ids = attr.ids_by_cpos(cpos_list)
for i, cpos in enumerate(cpos_list):
if cpos_attr_ids[i] != -1:
lookups['cpos_lookup'][cpos][attr.attrs['name']] = \
cpos_attr_ids[i]
occured_attr_ids = list(filter(lambda x: x != -1,
set(cpos_attr_ids)))
if not occured_attr_ids:
continue
subattrs = \
self.structural_attributes.list(filters={'part_of': attr})
if not subattrs:
continue
lookup_name = '{}_lookup'.format(attr.attrs['name'])
lookups[lookup_name] = {}
for attr_id in occured_attr_ids:
lookups[lookup_name][attr_id] = {}
for subattr in subattrs:
subattr_values = subattr.values_by_ids(occured_attr_ids)
for i, subattr_value in enumerate(subattr_values):
subattr_name = \
subattr.attrs['name'][(len(attr.attrs['name']) + 1):]
lookups[lookup_name][occured_attr_ids[i]][subattr_name] = \
subattr_value
return lookups
def drop(self):
return self.client.api.corpus_drop_corpus(self.attrs['api_name'])
def query(self, query, subcorpus_name='Results'):
return self.client.api.cqp_query(self.attrs['api_name'],
subcorpus_name, query)
@property
def alignment_attributes(self):
return AlignmentAttributeCollection(client=self.client, corpus=self)
@property
def positional_attributes(self):
return PositionalAttributeCollection(client=self.client, corpus=self)
@property
def structural_attributes(self):
return StructuralAttributeCollection(client=self.client, corpus=self)
@property
def subcorpora(self):
return SubcorpusCollection(client=self.client, corpus=self)
class CorpusCollection(Collection):
model = Corpus
def get(self, corpus_name):
return self.prepare_model(self.model._attrs(self.client, corpus_name))
def list(self):
return [self.prepare_model(self.model._attrs(self.client, corpus))
for corpus in self.client.api.corpus_list_coprora()]

78
app/corpora/cqi/models/ressource.py
View File

@ -1,78 +0,0 @@
class Model:
"""
A base class for representing a single object on the server.
"""
id_attribute = 'Id'
def __init__(self, attrs=None, client=None, collection=None):
#: A client pointing at the server that this object is on.
self.client = client
#: The collection that this model is part of.
self.collection = collection
#: The raw representation of this object from the API
self.attrs = attrs or {}
@staticmethod
def _attrs(client, key):
raise NotImplementedError
def __repr__(self):
return "<{}: {}>".format(self.__class__.__name__, self.id)
def __eq__(self, other):
return isinstance(other, self.__class__) and self.id == other.id
def __hash__(self):
return hash("{}:{}".format(self.__class__.__name__, self.id))
@property
def id(self):
"""
The ID of the object.
"""
return self.attrs.get(self.id_attribute)
def reload(self):
"""
Load this object from the server again and update ``attrs`` with the
new data.
"""
new_model = self.collection.get(self.id)
self.attrs = new_model.attrs
class Collection:
"""
A base class for representing all objects of a particular type on the
server.
"""
#: The type of object this collection represents, set by subclasses
model = None
def __init__(self, client=None):
#: The client pointing at the server that this collection of objects
#: is on.
self.client = client
def list(self):
raise NotImplementedError
def get(self, key):
raise NotImplementedError
def prepare_model(self, attrs):
"""
Create a model from a set of attributes.
"""
if isinstance(attrs, Model):
attrs.client = self.client
attrs.collection = self
return attrs
elif isinstance(attrs, dict):
return self.model(attrs=attrs, client=self.client, collection=self)
else:
raise Exception("Can't create {} from {}".format(
self.model.__name__, attrs))

103
app/corpora/cqi/models/subcorpora.py
View File

@ -1,103 +0,0 @@
from .ressource import Collection, Model
from ..api.specification import (CONST_FIELD_KEYWORD, CONST_FIELD_MATCH,
CONST_FIELD_MATCHEND, CONST_FIELD_TARGET)
class Subcorpus(Model):
id_attribute = 'api_name'
@staticmethod
def _attrs(client, corpus, name):
api_name = '{}:{}'.format(corpus.attrs['api_name'], name)
fields = {}
if client.api.cqp_subcorpus_has_field(api_name, CONST_FIELD_MATCH):
fields['match'] = CONST_FIELD_MATCH
if client.api.cqp_subcorpus_has_field(api_name,
CONST_FIELD_MATCHEND):
fields['matchend'] = CONST_FIELD_MATCHEND
if client.api.cqp_subcorpus_has_field(api_name, CONST_FIELD_TARGET):
fields['target'] = CONST_FIELD_TARGET
if client.api.cqp_subcorpus_has_field(api_name, CONST_FIELD_KEYWORD):
fields['keyword'] = CONST_FIELD_KEYWORD
return {'api_name': api_name,
'name': name,
'fields': fields,
'size': client.api.cqp_subcorpus_size(api_name)}
def drop(self):
return self.client.api.cqp_drop_subcorpus(self.attrs['api_name'])
def dump(self, field, first, last):
return self.client.api.cqp_dump_subcorpus(self.attrs['api_name'],
field, first, last)
def export(self, context=25, cutoff=float('inf'), expand_lists=False,
offset=0):
if self.attrs['size'] == 0:
return {"matches": []}
first_match = max(0, offset)
last_match = min((offset + cutoff - 1), (self.attrs['size'] - 1))
match_boundaries = zip(self.dump(self.attrs['fields']['match'],
first_match, last_match),
self.dump(self.attrs['fields']['matchend'],
first_match, last_match))
cpos_list = []
matches = []
for match_start, match_end in match_boundaries:
c = (match_start, match_end)
lc = rc = None
if context == 0:
cpos_list += list(range(match_start, (match_end + 1)))
else:
lc_rbound = max(0, (match_start - 1))
if lc_rbound != match_start:
lc_lbound = max(0, (match_start - context))
lc = (lc_lbound, lc_rbound)
cpos_list_lbound = lc_lbound
else:
cpos_list_lbound = match_start
rc_lbound = min((match_end + 1),
(self.collection.corpus.attrs['size'] - 1))
if rc_lbound != match_end:
rc_rbound = min((match_end + context),
(self.collection.corpus.attrs['size'] - 1))
rc = (rc_lbound, rc_rbound)
cpos_list_rbound = rc_rbound
else:
cpos_list_rbound = match_end
cpos_list += list(range(cpos_list_lbound,
(cpos_list_rbound + 1)))
if expand_lists:
match = {'lc': list(range(lc[0], (lc[1] + 1))),
'c': list(range(c[0], (c[1] + 1))),
'rc': list(range(rc[0], (rc[1] + 1)))}
else:
match = {'lc': lc, 'c': c, 'rc': rc}
matches.append(match)
lookups = self.collection.corpus.lookups_by_cpos(cpos_list)
return {'matches': matches, **lookups}
def fdist_1(self, cutoff, field, attribute):
return self.client.api.cqp_fdist_1(self.attrs['api_name'], cutoff,
field, attribute._name)
def fdist_2(self, cutoff, field_1, attribute_1, field_2, attribute_2):
return self.client.api.cqp_fdist_2(self.attrs['api_name'], cutoff,
field_1, attribute_1._name,
field_2, attribute_2._name)
class SubcorpusCollection(Collection):
model = Subcorpus
def __init__(self, client=None, corpus=None):
super(SubcorpusCollection, self).__init__(client=client)
self.corpus = corpus
def get(self, subcorpus_name):
return self.prepare_model(self.model._attrs(self.client, self.corpus,
subcorpus_name))
def list(self):
return [self.prepare_model(self.model._attrs(self.client, self.corpus, subcorpus)) # noqa
for subcorpus in self.client.api.cqp_list_subcorpora(self.corpus.attrs['api_name'])] # noqa

5
app/corpora/cqi/version.py
View File

@ -1,5 +0,0 @@
from .api.specification import MAJOR_VERSION, MINOR_VERSION
version = '{}.{}'.format(MAJOR_VERSION, MINOR_VERSION)
version_info = (MAJOR_VERSION, MINOR_VERSION)

2
app/corpora/events.py
View File

@ -1,11 +1,11 @@
from flask import current_app, request
from flask_login import current_user
from socket import gaierror
from . import cqi
from .. import db, socketio
from ..decorators import socketio_login_required
from ..events import connected_sessions
from ..models import Corpus, User
import cqi
import math

1
requirements.txt
View File

@ -1,3 +1,4 @@
cqi
email_validator
eventlet
Flask