Get problems from files in build_web_resources

The script now gets the list of languages, groups and problems from the
filesystem and inserts missing IDs in the database.

2 files changed, 122 insertions, 192 deletions

diff --git a/scripts/add_problem.py b/scripts/add_problem.py
deleted file mode 100755
index d8462cf..0000000
--- a/scripts/add_problem.py
+++ /dev/null
@@ -1,85 +0,0 @@
-#!/usr/bin/python3
-# coding=utf-8
-
-import db
-from .utils import filenamefy
-
-conn = db.get_connection()
-try:
-    cur = conn.cursor()
-    try:
-        new_lidentifier = None
-        # Get or add language.
-        cur.execute('select id, name, identifier from language order by id asc')
-        languages = cur.fetchall()
-        print('Languages:')
-        for lid, lname, lidentifier in languages:
-            print('  {}: {}'.format(lid, lname))
-
-        new_lid = input("Enter language ID or 'n' for new): ")
-        if new_lid == 'n':
-            new_lname = input('Enter name of the new language: ')
-            new_lidentifier = filenamefy(new_lname)
-            cur.execute('insert into language (name, identifier) values (%s, %s) returning id',
-                        (new_lname, new_lidentifier))
-            new_lid = cur.fetchone()[0]
-            print('Added new language "{}" with ID {} and identifier {}'.format(
-                new_lname, new_lid, new_lidentifier))
-        else:
-            new_lid = int(new_lid)
-            for lid, lname, lidentifier in languages:
-                print(lid, lname, lidentifier)
-                if lid == new_lid:
-                    new_lidentifier = lidentifier
-                    break
-        if new_lidentifier is None:
-            raise Exception('Language with ID {} does not exist'.format(new_lid))
-        print('Selected langauge {}'.format(new_lid))
-        print()
-
-        # Get or add problem group.
-        new_gidentifier = None
-        cur.execute('select id, name, identifier from problem_group order by id asc')
-        groups = cur.fetchall()
-        print('Problem groups:')
-        for gid, gname, gidentifier in groups:
-            print('  {}: {}'.format(gid, gname))
-        new_gid = input("Enter problem group ID or 'n' for new): ")
-        if new_gid == 'n':
-            new_gname = input('Enter name of the new problem group: ')
-            new_gidentifier = filenamefy(new_gname)
-            cur.execute('insert into problem_group (name, identifier) values (%s, %s) returning id',
-                        (new_gname, new_gidentifier))
-            new_gid = cur.fetchone()[0]
-            print('Added new problem group "{}" with ID {} and identifier {}'.format(
-                new_gname, new_gid, new_gidentifier))
-        else:
-            new_gid = int(new_gid)
-            for gid, gname, gidentifier in groups:
-                if gid == new_gid:
-                    new_gidentifier = gidentifier
-                    break
-        if new_gidentifier is None:
-            raise Exception('Group with ID {} does not exist'.format(new_gid))
-        print('Selected problem group {}'.format(new_gid))
-        print()
-
-        # Add problem.
-        new_pname = input('Enter name of the new problem: ')
-        new_pidentifier = filenamefy(new_pname)
-        cur.execute('insert into problem (language_id, problem_group_id, name, identifier, is_visible) values (%s, %s, %s, %s, %s) returning id',
-                    (new_lid, new_gid, new_pname, new_pidentifier, True))
-        new_pid = cur.fetchone()[0]
-        print('Added new problem "{}" with ID {} and identifier {}'.format(
-            new_pname, new_pid, new_pidentifier))
-        print('Data files should be placed in "{}/problems/{}/{}"'.format(
-            new_lidentifier, new_gidentifier, new_pidentifier))
-
-    finally:
-        cur.close()
-    conn.commit()
-except:
-    conn.rollback()
-    raise
-finally:
-    db.return_connection(conn)
diff --git a/scripts/build_web_resources.py b/scripts/build_web_resources.py
index 9a4ccec..e581fd0 100644
--- a/scripts/build_web_resources.py
+++ b/scripts/build_web_resources.py
@@ -1,8 +1,9 @@
 #!/usr/bin/python3
 # coding=utf-8
 
-"""This tool processes all problem files and database data, and outputs JSON
-files to describe problems, to be used as static web resources.
+"""This tool processes all problem files, and outputs JSON files to describe
+problems, to be used as static web resources. It also adds missing problem
+identifiers in the database.
 Before running the script define the following environment variables, if defaults are not okay:
 CODEQ_WEB_OUTPUT - directory where to write the output, defaults to /var/www/html/data
 CODEQ_PROBLEMS - directory where you have codeq-problems checked out, defaults to /var/local/codeq-problems
@@ -37,6 +38,7 @@ language_props = {  # for translation files inside the language subdirectory
     'hint': {}
 }
 language_common_props = {  # for common.py inside the language subdirectory
+    'id': None,  # database ID of the language
     'hint_type': {}  # type definitions of common hints
 }
 group_props = {  # for translation files inside the problem group subdirectory
@@ -44,6 +46,7 @@ group_props = {  # for translation files inside the problem group subdirectory
     'description': 'Description not set'
 }
 group_common_props = {  # for common.py inside the problem group subdirectory
+    'id': None,  # database ID of the problem group
     'number': 1  # display index of the problem group
 }
 problem_props = {  # for translation files inside the problem subdirectory
@@ -54,6 +57,7 @@ problem_props = {  # for translation files inside the problem subdirectory
     'hint': {}
 }
 problem_common_props = {  # for common.py inside the problem subdirectory
+    'id': None,  # database ID of the problem
     'number': 1,  # display index of problems inside their groups
     'visible': True,  # whether the problem is enabled (disabled problems are excluded from the application)
     'hint_type': {}  # type definitions of problem hints
@@ -65,9 +69,6 @@ groups = {}  # problem groups, info from database
 if not os.path.exists(output_path):
     os.mkdir(output_path)
 
-conn = db.get_connection()
-cur = conn.cursor()
-
 def load_translation_data(package, defaults):
     result = {}
     path = os.sep.join(package.split('.'))
@@ -138,121 +139,135 @@ def copy_web_resources(package, dst_dir_fragments):
                 shutil.copy(full_filename, dst_path)
                 node[filename] = True
 
-cur.execute('select id, name, identifier from language')
-row = cur.fetchone()
-while row:
-    languages[row[0]] = {'id': row[0], 'name': row[1], 'identifier': row[2]}
-    row = cur.fetchone()
-
-cur.execute('select id, name, identifier from problem_group')
-row = cur.fetchone()
-while row:
-    groups[row[0]] = {'id': row[0], 'name': row[1], 'identifier': row[2]}
-    row = cur.fetchone()
-
-cur.execute('select id, language_id, problem_group_id, identifier from problem where is_visible = true order by language_id, problem_group_id')
-previous_language_id = None
-previous_group_id = None
-lang_output_path = None
-lang_index = None
-row = cur.fetchone()
-
-def dump_language_defs():
-    if lang_index:
+def dump_language_defs(data, output_path):
+    if data:
         # sort groups and problems
-        groups = lang_index['groups']
+        groups = data['groups']
         for group in groups:
             group['problems'].sort(key=lambda p: p.get('number', 0))
         groups.sort(key=lambda p: p.get('number', 0))
         # write out the JSON file
-        with open(os.path.join(lang_output_path, 'language.json'), 'w') as f:
-            json.dump(lang_index, f, indent=2)
-
-while row:
-    # process language data, it all goes into the language directory language.json
-    try:
-        language_id = row[1]
-        if previous_language_id != language_id:
-            language = languages[language_id]
-            lang_identifier = language['identifier']
-            language_path = os.path.join(problems_path, lang_identifier)
-            if not (os.path.exists(language_path) and os.path.isdir(language_path)):
-                print('ERROR: the directory for language {0} does not exist: {1}'.format(lang_identifier, language_path))
-                continue
-            dump_language_defs()  # dump the previous language index
-            lang_output_path = os.path.join(output_path, lang_identifier)
-            if not os.path.exists(lang_output_path):
-                os.mkdir(lang_output_path)
-            previous_language_id = language_id
-            problem_groups_list = []
-            lang_index = load_common_data(lang_identifier, language_common_props)
-            lang_index['hint_type'] = process_hint_type(lang_index.get('hint_type', {}))  # process type definitions for common hints
-            lang_index['id'] = language_id
-            lang_index['identifier'] = lang_identifier
-            lang_index['groups'] = problem_groups_list
-            lang_index['translations'] = load_translation_data(lang_identifier, language_props)
-            previous_group_id = None
-            copy_web_resources(lang_identifier, [lang_identifier])
-
-        # process problem group data, it all goes into the language directory language.json
-        group_id = row[2]
-        if previous_group_id != group_id:
-            group = groups[group_id]
-            group_identifier = group['identifier']
-            group_path = os.path.join(language_path, 'problems', group_identifier)
-            if not (os.path.exists(group_path) and os.path.isdir(group_path)):
-                print('ERROR: the directory for group {0}/{1} does not exist: {2}'.format(lang_identifier, group_identifier, group_path))
+        with open(os.path.join(output_path, 'language.json'), 'w') as f:
+            json.dump(data, f, indent=2)
+
+conn = db.get_connection()
+cur = conn.cursor()
+
+def db_add(table, identifier, data):
+    cur.execute('select id from ' + table + ' where identifier = %s', (identifier,))
+    row = cur.fetchone()
+    if row is None:
+        data = sorted(data.items())
+        cols = [d[0] for d in data]
+        vals = [d[1] for d in data]
+        print('Inserting new {} in database: cols={} vals={}'.format(table, cols, vals))
+
+        args = ','.join(['%s'] * len(cols))
+        sql = 'insert into ' + table + ' (' + ','.join(cols) + ') values (' + args + ')'
+        cur.execute(sql, vals)
+
+try:
+    # get problem descriptors
+    for lang_identifier in os.listdir(problems_path):
+        lang_path = os.path.join(problems_path, lang_identifier)
+        if (not os.path.exists(os.path.join(lang_path, 'common.py')) or
+            not os.path.exists(os.path.join(lang_path, 'problems'))):
+            continue
+
+        # create language output directory
+        lang_output_path = os.path.join(output_path, lang_identifier)
+        if not os.path.exists(lang_output_path):
+            os.mkdir(lang_output_path)
+
+        lang_data = load_common_data(lang_identifier, language_common_props)
+        lang_data['hint_type'] = process_hint_type(lang_data.get('hint_type', {}))
+        lang_data['identifier'] = lang_identifier
+        lang_data['groups'] = []
+        lang_data['translations'] = load_translation_data(lang_identifier, language_props)
+        copy_web_resources(lang_identifier, [lang_identifier])
+        db_add('language', lang_identifier,
+                {'id': lang_data['id'], 'name': lang_identifier, 'identifier': lang_identifier})
+
+        groups_path = os.path.join(lang_path, 'problems')
+        for group_identifier in os.listdir(groups_path):
+            group_path = os.path.join(groups_path, group_identifier)
+            if not os.path.exists(os.path.join(group_path, 'common.py')):
                 continue
+
+            # create group directory
             group_output_path = os.path.join(lang_output_path, group_identifier)
             if not os.path.exists(group_output_path):
                 os.mkdir(group_output_path)
+
             group_package = lang_identifier + '.problems.' + group_identifier
-            previous_group_id = group_id
-            problems_list = []
             group_data = load_common_data(group_package, group_common_props)
-            group_data['id'] = group_id
             group_data['identifier'] = group_identifier
-            group_data['problems'] = problems_list
+            group_data['problems'] = []
             group_data['translations'] = load_translation_data(group_package, group_props)
-            problem_groups_list.append(group_data)
-            copy_web_resources(group_package, [lang_identifier, group_identifier])
-
-        # process problem data, from common.py goes into the language directory language.json, others go into problem subdirectory's problem.json
-        problem_id = row[0]
-        problem_identifier = row[3]
-        problem_path = os.path.join(group_path, problem_identifier)
-        if not (os.path.exists(problem_path) and os.path.isdir(problem_path)):
-            print('ERROR: the directory for problem {0}/{1}/{2} does not exist: {3}'.format(lang_identifier, group_identifier, problem_identifier, group_path))
-            continue
-        problem_package = group_package + '.' + problem_identifier
-        # load common data, for the language directory
-        common_data = load_common_data(problem_package, problem_common_props)
-        if not common_data['visible']:
-            continue  # problem is not visible, do not generate anything
-        del common_data['visible']  # we don't need this field in the GUI
-        hint_type = process_hint_type(common_data['hint_type'])  # save for later, to be used in problem_data below
-        del common_data['hint_type']  # we don't need this field in the language index
-        common_data['id'] = problem_id
-        common_data['identifier'] = problem_identifier
-        problems_list.append(common_data)
-        # load translations, and copy only problem names to the common data
-        problem_translations = load_translation_data(problem_package, problem_props)
-        name_translations = {}
-        for key, value in problem_translations.items():
-            name_translations[key] = {'name': value.get('name')}
-        common_data['translations'] = name_translations
-        # dump translations, for the problem subdirectory's problem.json
-        problem_data = {'id': problem_id, 'identifier': problem_identifier, 'translations': problem_translations, 'hint_type': hint_type}
-        problem_output_path = os.path.join(group_output_path, problem_identifier)
-        if not os.path.exists(problem_output_path):
-            os.mkdir(problem_output_path)
-        with open(os.path.join(problem_output_path, 'problem.json'), 'w') as f:
-            json.dump(problem_data, f, indent=2)
-        copy_web_resources(problem_package, [lang_identifier, group_identifier, problem_identifier])
-    finally:
-        row = cur.fetchone()
-
-dump_language_defs()  # dump the last language index
+            db_add('problem_group', group_identifier,
+                    {'id': group_data['id'], 'name': group_identifier, 'identifier': group_identifier})
+
+            for problem_identifier in os.listdir(group_path):
+                problem_path = os.path.join(group_path, problem_identifier)
+                if not os.path.exists(os.path.join(problem_path, 'common.py')):
+                    continue
+
+                problem_package = group_package + '.' + problem_identifier
+                common_data = load_common_data(problem_package, problem_common_props)
+                common_data['identifier'] = problem_identifier
+
+                if not common_data['visible']:
+                    continue  # problem is not visible, do not generate anything
+                del common_data['visible']  # we don't need this field in the GUI
+
+                # save for later, to be used in problem_data below
+                hint_type = process_hint_type(common_data['hint_type'])
+                del common_data['hint_type']  # we don't need this field in the language index
+
+                group_data['problems'].append(common_data)
+
+                # load translations, and copy only problem names to the common data
+                problem_translations = load_translation_data(problem_package, problem_props)
+                name_translations = {}
+                for key, value in problem_translations.items():
+                    name_translations[key] = {'name': value.get('name')}
+                common_data['translations'] = name_translations
+
+                problem_data = {
+                    'id': common_data['id'],
+                    'identifier': problem_identifier,
+                    'translations': problem_translations,
+                    'hint_type': hint_type
+                }
+
+                problem_output_path = os.path.join(group_output_path, problem_identifier)
+                if not os.path.exists(problem_output_path):
+                    os.mkdir(problem_output_path)
+                with open(os.path.join(problem_output_path, 'problem.json'), 'w') as f:
+                    json.dump(problem_data, f, indent=2)
+                copy_web_resources(problem_package, [lang_identifier, group_identifier, problem_identifier])
+                db_add('problem', problem_identifier, {
+                            'id': problem_data['id'],
+                            'language_id': lang_data['id'],
+                            'problem_group_id': group_data['id'],
+                            'name': problem_identifier,
+                            'identifier': problem_identifier
+                        })
+
+            # add only non-empty problem groups
+            if group_data['problems']:
+                copy_web_resources(group_package, [lang_identifier, group_identifier])
+                lang_data['groups'].append(group_data)
+
+        dump_language_defs(lang_data, lang_output_path)
+    cur.close()
+    conn.commit()
+except:
+    conn.rollback()
+    raise
+finally:
+    db.return_connection(conn)
+
 # dump the tree of resources
 with open(os.path.join(output_path, 'resources.json'), 'w') as f:
     json.dump(resource_tree, f, indent=2)