import requests working

.gitignore

@@ -35,3 +35,4 @@ mgiza-aligner/mgiza/mgizapp/src/mkcls/CMakeFiles/
 mgiza-aligner/mgiza/mgizapp/src/mkcls/Makefile
 mgiza-aligner/mgiza/mgizapp/src/mkcls/cmake_install.cmake
 __pycache__
+import-requests/request_handler.log

INSTALL.txt

@@ -88,3 +88,7 @@ mgiza-aligner:
 - cd mgiza, mgizapp
 - sudo apt-get install libboost-thread-dev
 - follow instructions in INSTALL
+
+import-requests:
+crontab -e
+* * * * * /path/to/handle_requests.py

cat/concordia_gate.php_pattern

@@ -2,9 +2,6 @@
 
 $url = 'http://@concordia_host@:@concordia_port@';
 $intervalsArray = array();
-foreach ($_POST["intervals"] as $interval) {
-    array_push($intervalsArray, [intval($interval[0]), intval($interval[1])]);
-}
 $data = array("operation" => $_POST["operation"],"tmId" => intval($_POST["tmId"]),"pattern" => $_POST["pattern"],"intervals" => $intervalsArray);
 
 // use key 'http' even if you send the request to https://...

cat/css/concordia_cat.css

@@ -320,7 +320,6 @@ th a.desc:after {
 }
 table tr td {
     line-height:200%;
-    max-width:200px;
 	padding: 6px;
 	text-align: left;
 	vertical-align: top;

cat/tm_manager.php_pattern

@@ -74,7 +74,14 @@ if ($_SERVER['REQUEST_METHOD'] == 'POST' ) {
     $errorMessage = addRequest($url, $_POST, $_FILES);
 }
 
-
+$statusesDictionary = array(
+    0 => "new",
+    1 => "processing - word alignment",
+    2 => "processing - adding to Concordia",
+    3 => "processing - regenerating index",
+    4 => "complete",
+    5 => "error"
+);
 
 $tmsData = postJson($url, array("operation" =>"getTmsInfo"));
 $requestsData = postJson($url, array("operation" =>"getRequestsInfo"));
@@ -149,9 +156,9 @@ $languagesData = postJson($url, array("operation" =>"getLanguages"));
                 ?>
                 </select>
 
-                <table>
+                <table style="table-layout:fixed">
                     <tr>
-                        <td width="70%">
+                        <td style="min-width:500px">
                             <label for="src_file">Source file (TXT)</label>
                             <input id="src_file" name="src_file" type="file">
                         </td>
@@ -226,7 +233,7 @@ $languagesData = postJson($url, array("operation" =>"getLanguages"));
                         }
                         ?>
                     </td>
-                    <td><?= $request->status ?></td>
+                    <td><?= $statusesDictionary[$request->status] ?></td>
                     <td>
                         <?php
                         if($request->type == 0) {

db/concordia_server.sql

@@ -48,3 +48,5 @@ CREATE TABLE alignment (
 );
 
 CREATE INDEX ON alignment(unit_id, source_token_pos);
+
+CREATE INDEX ON unit(tm_id);

import-requests/handle_requests.py

@@ -1,12 +1,10 @@
 #!/usr/bin/python3
 # -*- coding: utf-8 -*-
 
-import psycopg2, os, shutil, subprocess, urllib3, time, json
+import psycopg2, os, sys, shutil, subprocess, urllib3, time, json, logging, errno
 
 from importlib.machinery import SourceFileLoader
 
-BUFFER_SIZE = 500
-
 def postJson(address, data):
     http = urllib3.PoolManager()
     response = http.request('POST', address, headers={'Content-Type': 'application/json'},body=json.dumps(data).encode('utf-8'))
@@ -26,25 +24,140 @@ def file_len(fname):
     return i + 1
 
 
-mgiza_path = os.path.dirname(os.path.realpath(__file__))+'/../mgiza-aligner'
+def pid_exists(pid):
+    """Check whether pid exists in the current process table.
+    UNIX only.
+    """
+    if pid < 0:
+        return False
+    if pid == 0:
+        # According to "man 2 kill" PID 0 refers to every process
+        # in the process group of the calling process.
+        # On certain systems 0 is a valid PID but we have no way
+        # to know that in a portable fashion.
+        raise ValueError('invalid PID 0')
+    try:
+        os.kill(pid, 0)
+    except OSError as err:
+        if err.errno == errno.ESRCH:
+            # ESRCH == No such process
+            return False
+        elif err.errno == errno.EPERM:
+            # EPERM clearly means there's a process to deny access to
+            return True
+        else:
+            # According to "man 2 kill" possible error values are
+            # (EINVAL, EPERM, ESRCH)
+            raise
+    else:
+        return True
 
-conn = psycopg2.connect("dbname='concordia_server' user='concordia' host='localhost' port='6543' password='concordia'")
+def updateRequestStatus(connection_string, request_id, new_status):
+    conn = psycopg2.connect(connection_string)
     cur = conn.cursor()
-cur.execute("""select request.id, request.source_file_path, request.target_file_path, request.name, src_lang.id as src_lang_id, src_lang.code as src_code, trg_lang.id as trg_lang_id, trg_lang.code as trg_code, request.status, request.type, request.tm_id from request inner join language as src_lang on src_lang.id = request.source_lang_id inner join language as trg_lang on trg_lang.id = request.target_lang_id order by request.created limit 1""")
+    cur.execute("UPDATE request SET status=%s WHERE id=%s", (new_status, request_id))
+    conn.commit()
+    cur.close()
+    conn.close()
 
+def getPendingRequest(connection_string):
+    conn = psycopg2.connect(connection_string)
+    cur = conn.cursor()
+    cur.execute("""select request.id, request.source_file_path, request.target_file_path, request.name, src_lang.id as src_lang_id, src_lang.code as src_code, trg_lang.id as trg_lang_id, trg_lang.code as trg_code, request.status, request.type, request.tm_id from request inner join language as src_lang on src_lang.id = request.source_lang_id inner join language as trg_lang on trg_lang.id = request.target_lang_id where request.status = 0 order by request.created limit 1""")
 
     request = cur.fetchone()
-request_id, src_file_path, trg_file_path, tm_name, src_lang_id, src_lang_code, trg_lang_id, trg_lang_code, status, tm_type, tm_id  = request
+    cur.close()
-request_corpus_path = mgiza_path+'/corpora/request_'+str(request_id)
+    conn.close()
-os.makedirs(request_corpus_path)
+    return request
-shutil.copy(src_file_path, request_corpus_path+'/src.txt')
-shutil.copy(trg_file_path, request_corpus_path+'/trg.txt')
 
-subprocess.run(["make","SRC_LANG="+src_lang_code, "TRG_LANG="+trg_lang_code, "CORPUS_NAME=request_"+str(request_id)], cwd=mgiza_path)
+def retrieve_and_append(connection_string, tm_id, src_file_path, trg_file_path):
+    conn = psycopg2.connect(connection_string)
+    cur = conn.cursor()
+    cur.execute("SELECT source_segment, target_segment FROM unit WHERE tm_id=%d" % tm_id)
+    with open(src_file_path, 'a') as src_file, open(trg_file_path, 'a') as trg_file:
+        for row in cur:
+            src_file.write(row[0]+'\n')
+            trg_file.write(row[1]+'\n')
 
     cur.close()
     conn.close()
 
+def clearTm(connection_string, tm_id):
+    conn = psycopg2.connect(connection_string)
+    cur = conn.cursor()
+    cur.execute("DELETE FROM unit WHERE tm_id=%d" %tm_id)
+    conn.commit()
+    cur.close()
+    conn.close()
+
+def getLangsFromTM(connection_string, tm_id):
+    conn = psycopg2.connect(connection_string)
+    cur = conn.cursor()
+    cur.execute("select src_lang.code, trg_lang.code from tm inner join language as src_lang on tm.source_lang_id = src_lang.id inner join language as trg_lang on tm.target_lang_id = trg_lang.id where tm.id = %d" % tm_id)
+    langs = cur.fetchone()
+    cur.close()
+    conn.close()
+    return langs
+
+# ----------- main -------------------------------------------------------------
+
+BUFFER_SIZE = 500
+
+src_dir = os.path.dirname(os.path.realpath(__file__))
+
+logger = logging.getLogger('request_handler')
+hdlr = logging.FileHandler(src_dir+'/request_handler.log')
+formatter = logging.Formatter('%(asctime)s %(levelname)s %(message)s')
+hdlr.setFormatter(formatter)
+logger.addHandler(hdlr)
+logger.setLevel(logging.INFO)
+
+
+pidfile_path = src_dir+'/request_handler.pid'
+if os.path.isfile(pidfile_path):
+    with open(pidfile_path) as pidfile:
+        pid = int(pidfile.readline().strip())
+        logger.info('Found pid file with pid: %d' % pid)
+        if pid_exists(pid):
+            logger.info('The process with pid=%d is actually running, exiting' % pid)
+            sys.exit(0)
+        else:
+            logger.warning('The process with pid=%d is not running anymore. Ignoring the pid file.' % pid)
+
+with open(pidfile_path, 'w') as pidfile:
+    pidfile.write(str(os.getpid()))
+
+mgiza_path = src_dir+'/../mgiza-aligner'
+
+connection_string = "dbname='concordia_server' user='concordia' host='localhost' port='6543' password='concordia'"
+
+try:
+    request = getPendingRequest(connection_string)
+    if request is not None:
+        request_id, src_file_path, trg_file_path, tm_name, src_lang_id, src_lang_code, trg_lang_id, trg_lang_code, status, tm_type, tm_id  = request
+
+        if tm_type == 1:
+            src_lang_code, trg_lang_code = getLangsFromTM(connection_string, tm_id)
+
+        logger.info("Working on request: %d. Starting the mgiza aligner." % request_id)
+        updateRequestStatus(connection_string, request_id, 1)
+
+        request_corpus_path = mgiza_path+'/corpora/request_'+str(request_id)
+        os.makedirs(request_corpus_path)
+
+        shutil.copy(src_file_path, request_corpus_path+'/src.txt')
+        shutil.copy(trg_file_path, request_corpus_path+'/trg.txt')
+
+        if tm_type == 1:
+            # extending existing TM. We must retrieve existing units from the db.
+            logger.info("This is an extend request. Appending units from the tm %d to the src and trg files."%tm_id)
+            retrieve_and_append(connection_string, tm_id, request_corpus_path+'/src.txt', request_corpus_path+'/trg.txt')
+
+        subprocess.run(["make","SRC_LANG="+src_lang_code, "TRG_LANG="+trg_lang_code, "CORPUS_NAME=request_"+str(request_id)], cwd=mgiza_path, check=True, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+
+        logger.info("Mgiza alignment complete, starting adding the TM to Concordia.")
+        updateRequestStatus(connection_string, request_id, 2)
+
         host = SourceFileLoader("host", os.path.dirname(os.path.realpath(__file__))+'/../tests/host.py').load_module()
         address = 'http://'+host.concordia_host
         if len(host.concordia_port) > 0:
@@ -61,6 +174,8 @@ if (file_len(sourceFile) != file_len(targetFile)):
         if (file_len(alignmentsFile) != 3*file_len(sourceFile)):
             raise Exception("alignments file is not exactly 3 times longer than source and target")
 
+        if tm_type == 0:
+            # create new TM
             data = {
                 'operation': 'addTm',
                 'sourceLangId':src_lang_id,
@@ -72,7 +187,10 @@ data = {
             response = postJson(address, data)
 
             tmId = int(response['newTmId'])
-print("Added new tm: %d" % tmId)
+            logger.info("Added new tm: %d" % tmId)
+        else:
+            clearTm(connection_string, tm_id)
+            tmId = tm_id
 
         data = {
             'operation': 'addAlignedLemmatizedSentences',
@@ -102,7 +220,8 @@ if len(examples) > 0:
             data['examples'] = examples
             add_examples(address, data)
 
-print("Generating index...")
+        updateRequestStatus(connection_string, request_id, 3)
+        logger.info("Generating index...")
         start = time.time()
         data = {
             'operation': 'refreshIndex',
@@ -112,4 +231,12 @@ data = {
         response = postJson(address, data)
 
         end = time.time()
-print("Index regeneration complete. The operation took %.4f s" % (end - start))
+        logger.info("Index regeneration complete. The operation took %.4f s" % (end - start))
+
+except Exception as e:
+    logger.error("Error occured: "+str(e))
+    updateRequestStatus(connection_string, request_id, 5)
+else:
+    updateRequestStatus(connection_string, request_id, 4)
+finally:
+    os.remove(pidfile_path)

mgiza-aligner/Makefile

@@ -66,11 +66,11 @@ corpora/$(CORPUS_NAME)/src_clean.lem: corpora/$(CORPUS_NAME)/src_clean.tok
 
 
 corpora/$(CORPUS_NAME)/trg.tok: corpora/$(CORPUS_NAME)/trg.txt
-	concordia-sentence-tokenizer -c ../concordia.cfg < $< > $@
+	/usr/local/bin/concordia-sentence-tokenizer -c ../concordia.cfg < $< > $@
 
 
 corpora/$(CORPUS_NAME)/src.tok: corpora/$(CORPUS_NAME)/src.txt
-	concordia-sentence-tokenizer -c ../concordia.cfg < $< > $@
+	/usr/local/bin/concordia-sentence-tokenizer -c ../concordia.cfg < $< > $@
 
 
 corpora/$(CORPUS_NAME)/src_clean.txt corpora/$(CORPUS_NAME)/trg_clean.txt corpora/$(CORPUS_NAME)/src_clean.tok corpora/$(CORPUS_NAME)/trg_clean.tok: corpora/$(CORPUS_NAME)/src.txt corpora/$(CORPUS_NAME)/trg.txt corpora/$(CORPUS_NAME)/src.tok corpora/$(CORPUS_NAME)/trg.tok