kaqing 2.0.145__py3-none-any.whl → 2.0.189__py3-none-any.whl

adam/commands/export/exporter.py

@@ -1,23 +1,20 @@
-from concurrent.futures import ThreadPoolExecutor, as_completed
 from datetime import datetime
-import functools
-import re
 import time
-import traceback
 
-from adam.commands.cql.cql_utils import cassandra_table_names, run_cql, table_spec
-from adam.commands.export.export_databases import ExportDatabases
+from adam.commands.command import InvalidArgumentsException
+from adam.commands.cql.utils_cql import cassandra_table_names, run_cql, table_spec
+from adam.commands.export.export_databases import export_db
+from adam.commands.export.export_sessions import ExportSessions
 from adam.commands.export.importer import Importer
 from adam.commands.export.importer_athena import AthenaImporter
 from adam.commands.export.importer_sqlite import SqliteImporter
-from adam.commands.export.utils_export import ExportSpec, ExportTableStatus, ExportTableSpec, ImportSpec, csv_dir, find_files
+from adam.commands.export.utils_export import ExportSpec, ExportTableStatus, ExportTableSpec, ImportSpec, csv_dir, find_files, state_with_pod
 from adam.config import Config
 from adam.pod_exec_result import PodExecResult
 from adam.repl_state import ReplState
-from adam.utils import elapsed_time, log2, ing
+from adam.utils import debug, log, parallelize, log2, ing, log_exc
 from adam.utils_k8s.cassandra_nodes import CassandraNodes
 from adam.utils_k8s.pods import log_prefix
-from adam.utils_k8s.statefulsets import StatefulSets
 
 class Exporter:
     def export_tables(args: list[str], state: ReplState, export_only: bool = False, max_workers = 0) -> tuple[list[str], ExportSpec]:
@@ -25,7 +22,7 @@ class Exporter:
             log2('export-only for testing')
 
         spec: ExportSpec = None
-        try:
+        with log_exc(True):
             spec = Exporter.export_spec(' '.join(args), state)
 
             statuses, spec = Exporter._export_tables(spec, state, max_workers=max_workers, export_state='init')
@@ -33,8 +30,6 @@ class Exporter:
                 return statuses, spec
 
             return Exporter._export_tables(spec, state, export_only, max_workers, 'pending_export')
-        except Exception as e:
-            log2(e)
 
         return [], None
 
@@ -53,12 +48,19 @@ class Exporter:
                         raise Exception(f"You're currently using {importer_from_session} export database. You cannot export tables with {spec.importer} type database.")
             else:
                 spec.importer = Importer.importer_from_session(session)
+
+            if spec.importer == 'athena' and not AthenaImporter.ping():
+                raise Exception('Credentials for Athena is not present.')
         else:
             if not spec.importer:
                 spec.importer = Config().get('export.default-importer', 'sqlite')
 
             prefix = Importer.prefix_from_importer(spec.importer)
             session = f'{prefix}{datetime.now().strftime("%Y%m%d%H%M%S")[3:]}'
+
+            if spec.importer == 'athena' and not AthenaImporter.ping():
+                raise Exception('Credentials for Athena is not present.')
+
             if spec.importer != 'csv':
                 state.export_session = session
 
@@ -66,10 +68,10 @@ class Exporter:
 
         return spec
 
-    def import_session(args: list[str], state: ReplState, max_workers = 0) -> tuple[list[str], ExportSpec]:
+    def import_session(spec_str: str, state: ReplState, max_workers = 0) -> tuple[list[str], ExportSpec]:
         import_spec: ImportSpec = None
-        try:
-            import_spec = Exporter.import_spec(' '.join(args), state)
+        with log_exc(True):
+            import_spec = Exporter.import_spec(spec_str, state)
             tables, status_in_whole = ExportTableStatus.from_session(state.sts, state.pod, state.namespace, import_spec.session)
             if status_in_whole == 'done':
                 log2(f'The session has been completely done - no more csv files are found.')
@@ -77,17 +79,34 @@ class Exporter:
 
             spec = ExportSpec(None, None, importer=import_spec.importer, tables=[ExportTableSpec.from_status(table) for table in tables], session=import_spec.session)
 
-            return Exporter._export_tables(spec, state, max_workers=max_workers)
-        except Exception as e:
-            if Config().is_debug():
-                traceback.print_exception(e)
-            else:
-                log2(e)
+            return Exporter._export_tables(spec, state, max_workers=max_workers, export_state = 'import')
+
+        return [], None
+
+    def import_local_csv_files(spec_str: str, state: ReplState, max_workers = 0) -> tuple[list[str], ExportSpec]:
+        spec: ImportSpec = None
+        with log_exc(True):
+            spec = Exporter.import_spec(spec_str, state, files=True)
+            if not spec.table_name:
+                log2(f"Use 'as <database-name>.<table-name>'.")
+                raise InvalidArgumentsException()
+
+            d_t = spec.table_name.split('.')
+            if len(d_t) != 2:
+                log2(f'Need <database-name>.<table-name> format for target table.')
+                raise InvalidArgumentsException()
+
+            database = d_t[0]
+            table = d_t[1]
+            im = AthenaImporter() if spec.importer == 'athena' else SqliteImporter()
+
+            with parallelize(spec.files, max_workers, msg='Importing|Imported {size} csv files') as exec:
+                return exec.map(lambda f: im.import_from_local_csv(state, database, table, f, len(spec.files) > 1, True)), spec
 
         return [], None
 
-    def import_spec(spec_str: str, state: ReplState):
-        spec: ImportSpec = ImportSpec.parse_specs(spec_str)
+    def import_spec(spec_str: str, state: ReplState, files = False):
+        spec: ImportSpec = ImportSpec.parse_specs(spec_str, files=files)
 
         session = state.export_session
         if session:
@@ -99,19 +118,26 @@ class Exporter:
                 spec.importer = Importer.importer_from_session(state.export_session)
                 if not spec.importer:
                     spec.importer = Config().get('export.default-importer', 'sqlite')
+
+            if spec.importer == 'athena' and not AthenaImporter.ping():
+                raise Exception('Credentials for Athena is not present.')
         else:
-            if spec.importer:
-                if not AthenaImporter.ping():
-                    raise Exception('Credentials for Athena are not present.')
-            else:
+            if not spec.importer:
                 spec.importer = Importer.importer_from_session(spec.session)
 
-            if spec.importer == 'csv':
+            if not spec.importer or spec.importer == 'csv':
                 spec.importer = Config().get('export.default-importer', 'sqlite')
 
+            if spec.importer == 'athena' and not AthenaImporter.ping():
+                raise Exception('Credentials for Athena is not present.')
+
             prefix = Importer.prefix_from_importer(spec.importer)
-            session = f'{prefix}{spec.session[1:]}'
-            state.export_session = session
+            if spec.session:
+                spec.session = f'{prefix}{spec.session[1:]}'
+            else:
+                spec.session = f'{prefix}{datetime.now().strftime("%Y%m%d%H%M%S")[3:]}'
+
+            state.export_session = spec.session
 
         return spec
 
@@ -128,20 +154,14 @@ class Exporter:
         if export_state == 'init':
             CassandraNodes.exec(state.pod, state.namespace, f'rm -rf {csv_dir()}/{spec.session}_*', show_out=Config().is_debug(), shell='bash')
 
-        if max_workers > 1 and len(spec.tables) > 1:
-            log2(f'Executing on {len(spec.tables)} Cassandra tables in parallel...')
-            start_time = time.time()
-            try:
-                with ThreadPoolExecutor(max_workers=max_workers) as executor:
-                    futures = [executor.submit(Exporter.export_table, table, state, spec.session, spec.importer, export_only, True, consistency=spec.consistency, export_state=export_state) for table in spec.tables]
-                    if len(futures) == 0:
-                        return [], spec
-
-                return [future.result() for future in as_completed(futures)], spec
-            finally:
-                log2(f"{len(spec.tables)} parallel table export elapsed time: {elapsed_time(start_time)} with {max_workers} workers")
-        else:
-            return [Exporter.export_table(table, state, spec.session, spec.importer, export_only, multi_tables=len(spec.tables) > 1, consistency=spec.consistency, export_state=export_state) for table in spec.tables], spec
+        action = f'[{spec.session}] Exporting|Exported'
+        if export_state == 'init':
+            action = f'[{spec.session}] Preparing|Prepared'
+        elif export_state == 'import':
+            action = f'[{spec.session}] Importing|Imported'
+
+        with parallelize(spec.tables, max_workers, msg=action + ' {size} Cassandra tables') as exec:
+            return exec.map(lambda table: Exporter.export_table(table, state, spec.session, spec.importer, export_only, len(spec.tables) > 1, consistency=spec.consistency, export_state=export_state)), spec
 
     def export_table(spec: ExportTableSpec, state: ReplState, session: str, importer: str, export_only = False, multi_tables = True, consistency: str = None, export_state=None):
         s: str = None
@@ -167,8 +187,7 @@ class Exporter:
             status: ExportTableStatus = ExportTableStatus.from_log_file(state.pod, state.namespace, session, log_file)
             while status.status != 'done':
                 if status.status == 'export_in_pregress':
-                    if Config().is_debug():
-                        log2('Exporting to CSV is still in progess, sleeping for 1 sec...')
+                    debug('Exporting to CSV is still in progess, sleeping for 1 sec...')
                     time.sleep(1)
                 elif status.status == 'exported':
                     log_file = Exporter.rename_to_pending_import(spec, state, session, target_table)
@@ -202,7 +221,7 @@ class Exporter:
         queries.append(f"COPY {spec.keyspace}.{table}({columns}) TO '{csv_file}' WITH HEADER = TRUE")
         r: PodExecResult = ing(
             f'[{session}] Dumping table {spec.keyspace}.{table}{f" with consistency {consistency}" if consistency else ""}',
-            lambda: run_cql(state, ';'.join(queries), show_out=Config().is_debug(), background=True, log_file=log_file),
+            lambda: run_cql(state, ';'.join(queries), show_out=Config().is_debug(), backgrounded=True, log_file=log_file),
             suppress_log=suppress_ing_log)
 
         return log_file
@@ -217,114 +236,7 @@ class Exporter:
 
     def import_from_csv(spec: ExportTableSpec, state: ReplState, session: str, importer: str, table: str, target_table: str, columns: str, multi_tables = True, create_db = False):
         im = AthenaImporter() if importer == 'athena' else SqliteImporter()
-        return im.import_from_csv(state.pod, state.namespace, state.export_session, session if session else state.export_session, spec.keyspace, table, target_table, columns, multi_tables, create_db)
-
-    def clear_export_session_cache():
-        Exporter.find_export_sessions.cache_clear()
-        Exporter.export_session_names.cache_clear()
-
-    @functools.lru_cache()
-    def export_session_names(sts: str, pod: str, namespace: str, importer: str = None, export_state = None):
-        if not sts or not namespace:
-            return []
-
-        if not pod:
-            pod = StatefulSets.pod_names(sts, namespace)[0]
-
-        if not pod:
-            return []
-
-        return [session for session, state in Exporter.find_export_sessions(pod, namespace, importer).items() if not export_state or state == export_state]
-
-    @functools.lru_cache()
-    def find_export_sessions(pod: str, namespace: str, importer: str = None, limit = 100):
-        sessions: dict[str, str] = {}
-
-        prefix = Importer.prefix_from_importer(importer)
-
-        log_files: list[str] = find_files(pod, namespace, f'{log_prefix()}-{prefix}*_*.log*')
-
-        if not log_files:
-            return {}
-
-        for log_file in log_files[:limit]:
-            m = re.match(f'{log_prefix()}-(.*?)_.*\.log?(.*)', log_file)
-            if m:
-                s = m.group(1)
-                state = m.group(2) # '', '.pending_import', '.done'
-                if state:
-                    state = state.strip('.')
-                else:
-                    state = 'in_export'
-
-                if s not in sessions:
-                    sessions[s] = state
-                elif sessions[s] == 'done' and state != 'done':
-                    sessions[s] = state
-
-        return sessions
-
-    def clean_up_all_sessions(sts: str, pod: str, namespace: str):
-        if not sts or not namespace:
-            return False
-
-        if not pod:
-            pod = StatefulSets.pod_names(sts, namespace)[0]
-
-        CassandraNodes.exec(pod, namespace, f'rm -rf {csv_dir()}/*', show_out=Config().is_debug(), shell='bash')
-        CassandraNodes.exec(pod, namespace, f'rm -rf {log_prefix()}-*.log*', show_out=Config().is_debug(), shell='bash')
-
-        return True
-
-    def clean_up_sessions(sts: str, pod: str, namespace: str, sessions: list[str], max_workers = 0):
-        if not sessions:
-            return []
-
-        if not max_workers:
-            max_workers = Config().action_workers('export', 8)
-
-        if max_workers > 1 and len(sessions) > 1:
-            log2(f'Executing on {len(sessions)} export session clean ups in parallel...')
-            start_time = time.time()
-            try:
-                with ThreadPoolExecutor(max_workers=max_workers) as executor:
-                    futures = [executor.submit(Exporter.clean_up_session, sts, pod, namespace, session, True) for session in sessions]
-                    if len(futures) == 0:
-                        return []
-
-                return [future.result() for future in as_completed(futures)]
-            finally:
-                log2(f"{len(sessions)} parallel session clean ups elapsed time: {elapsed_time(start_time)} with {max_workers} workers")
-        else:
-            return [Exporter.clean_up_session(sts, pod, namespace, session) for session in sessions]
-
-    def clean_up_session(sts: str, pod: str, namespace: str, session: str, multi_tables = True):
-        if not sts or not namespace:
-            return 0, 0
-
-        if not pod:
-            pod = StatefulSets.pod_names(sts, namespace)[0]
-
-        if not pod:
-            return 0, 0
-
-        csv_cnt = 0
-        log_cnt = 0
-
-        log_files: list[str] = find_files(pod, namespace, f'{log_prefix()}-{session}_*.log*')
-
-        for log_file in log_files:
-            m = re.match(f'{log_prefix()}-{session}_(.*?)\.(.*?)\.log.*', log_file)
-            if m:
-                table = m.group(2)
-
-                CassandraNodes.exec(pod, namespace, f'rm -rf {csv_dir()}/{session}_{table}', show_out=not multi_tables, shell='bash')
-                csv_cnt += 1
-
-                CassandraNodes.exec(pod, namespace, f'rm -rf {log_file}', show_out=not multi_tables, shell='bash')
-                log_cnt += 1
-
-        return csv_cnt, log_cnt
+        return im.import_from_csv(state, session if session else state.export_session, spec.keyspace, table, target_table, columns, multi_tables, create_db)
 
     def resove_table_n_columns(spec: ExportTableSpec, state: ReplState, include_ks_in_target = False, importer = 'sqlite'):
         table = spec.table
@@ -350,15 +262,71 @@ class Exporter:
 
         return table, target_table, columns
 
-    def drop_databases(sts: str, pod: str, namespace: str, db: str = None):
-        importer = None
-        if db:
-            importer = Importer.importer_from_session(db)
-
-        sessions_done = Exporter.export_session_names(sts, pod, namespace, importer=importer, export_state='done')
-        sessions = ExportDatabases.sessions_from_dbs(ExportDatabases.drop_export_dbs(db))
-        if sessions_done and sessions:
-            intersects = list(set(sessions_done) & set(sessions))
-            with ing(f'Cleaning up {len(intersects)} completed sessions'):
-                Exporter.clean_up_sessions(sts, pod, namespace, list(intersects))
-                Exporter.clear_export_session_cache()
+class ExportService:
+    def __init__(self, handler: 'ExporterHandler'):
+        self.handler = handler
+
+    def export(self, args: list[str], export_only=False):
+        state = self.handler.state
+        export_session = state.export_session
+        spec: ExportSpec = None
+        try:
+            with state_with_pod(state) as state:
+                # --export-only for testing only
+                statuses, spec = Exporter.export_tables(args, state, export_only=export_only)
+                if not statuses:
+                    return state
+
+                ExportSessions.clear_export_session_cache()
+
+                if spec.importer == 'csv' or export_only:
+                    ExportSessions.show_session(state.sts, state.pod, state.namespace, spec.session)
+                else:
+                    log()
+                    with export_db(state) as dbs:
+                        dbs.show_database()
+        finally:
+            # if exporting to csv, do not bind the new session id to repl state
+            if spec and spec.importer == 'csv':
+                state.export_session = export_session
+
+        return state
+
+    def import_session(self, spec_str: str):
+        state = self.handler.state
+
+        tables, _ = Exporter.import_session(spec_str, state)
+        if tables:
+            ExportSessions.clear_export_session_cache()
+
+            log()
+            with export_db(state) as dbs:
+                dbs.show_database()
+
+        return state
+
+    def import_files(self, spec_str: str):
+        state = self.handler.state
+
+        tables, _ = Exporter.import_local_csv_files(spec_str, state)
+        if tables:
+            ExportSessions.clear_export_session_cache()
+
+            log()
+            with export_db(state) as dbs:
+                dbs.show_database()
+
+        return state
+
+class ExporterHandler:
+    def __init__(self, state: ReplState):
+        self.state = state
+
+    def __enter__(self):
+        return ExportService(self)
+
+    def __exit__(self, exc_type, exc_val, exc_tb):
+        return False
+
+def export(state: ReplState):
+    return ExporterHandler(state)

adam/commands/export/import_files.py

@@ -0,0 +1,44 @@
+from adam.commands import validate_args
+from adam.commands.command import Command
+from adam.commands.export.export_sessions import ExportSessions
+from adam.commands.export.exporter import export
+from adam.commands.export.utils_export import state_with_pod
+from adam.repl_state import ReplState, RequiredState
+
+class ImportCSVFiles(Command):
+    COMMAND = 'import files'
+
+    # the singleton pattern
+    def __new__(cls, *args, **kwargs):
+        if not hasattr(cls, 'instance'): cls.instance = super(ImportCSVFiles, cls).__new__(cls)
+
+        return cls.instance
+
+    def __init__(self, successor: Command=None):
+        super().__init__(successor)
+
+    def command(self):
+        return ImportCSVFiles.COMMAND
+
+    def required(self):
+        return RequiredState.CLUSTER_OR_POD
+
+    def run(self, cmd: str, state: ReplState):
+        if not(args := self.args(cmd)):
+            return super().run(cmd, state)
+
+        with self.validate(args, state) as (args, state):
+            with validate_args(args, state, name='file') as spec:
+                with state_with_pod(state) as state:
+                    with export(state) as exporter:
+                        return exporter.import_files(spec)
+
+    def completion(self, state: ReplState):
+        # warm up cache
+        # ExportSessions.export_session_names(state.sts, state.pod, state.namespace)
+        # ExportSessions.export_session_names(state.sts, state.pod, state.namespace, export_state='pending_import')
+
+        return {}
+
+    def help(self, _: ReplState):
+        return f'{ImportCSVFiles.COMMAND} <file-names,...>\t import files in session to Athena or SQLite'

adam/commands/export/import_session.py

@@ -1,9 +1,9 @@
+from adam.commands import validate_args
 from adam.commands.command import Command
-from adam.commands.export.export_databases import ExportDatabases
-from adam.commands.export.exporter import Exporter
+from adam.commands.export.export_sessions import ExportSessions
+from adam.commands.export.exporter import export
+from adam.commands.export.utils_export import state_with_pod
 from adam.repl_state import ReplState, RequiredState
-from adam.utils import log, log2
-from adam.utils_k8s.statefulsets import StatefulSets
 
 class ImportSession(Command):
     COMMAND = 'import session'
@@ -27,40 +27,16 @@ class ImportSession(Command):
         if not(args := self.args(cmd)):
             return super().run(cmd, state)
 
-        state, args = self.apply_state(args, state)
-        if not self.validate_state(state):
-            return state
-
-        if not args:
-            if state.in_repl:
-                log2('Specify export session name.')
-            else:
-                log2('* Export session name is missing.')
-
-                Command.display_help()
-
-            return 'command-missing'
-
-        if not state.pod:
-            state.push()
-            state.pod = StatefulSets.pod_names(state.sts, state.namespace)[0]
-
-        try:
-            tables, _ = Exporter.import_session(args, state)
-            if tables:
-                Exporter.clear_export_session_cache()
-
-                log()
-                ExportDatabases.display_export_db(state.export_session)
-        finally:
-            state.pop()
-
-        return state
+        with self.validate(args, state) as (args, state):
+            with validate_args(args, state, name='export session') as spec:
+                with state_with_pod(state) as state:
+                    with export(state) as exporter:
+                        return exporter.import_session(spec)
 
     def completion(self, state: ReplState):
         # warm up cache
-        Exporter.export_session_names(state.sts, state.pod, state.namespace)
-        Exporter.export_session_names(state.sts, state.pod, state.namespace, export_state='pending_import')
+        # ExportSessions.export_session_names(state.sts, state.pod, state.namespace)
+        # ExportSessions.export_session_names(state.sts, state.pod, state.namespace, export_state='pending_import')
 
         return {}
 

adam/commands/export/importer.py

@@ -2,6 +2,7 @@ from abc import abstractmethod
 
 from adam.commands.export.utils_export import csv_dir
 from adam.config import Config
+from adam.repl_state import ReplState
 from adam.utils import ing
 from adam.utils_k8s.cassandra_nodes import CassandraNodes
 from adam.utils_k8s.pods import log_prefix
@@ -12,10 +13,20 @@ class Importer:
         pass
 
     @abstractmethod
-    def import_from_csv(self, pod: str, namespace: str, to_session: str, from_session: str, keyspace: str, table: str, target_table: str, columns: str, multi_tables = True, create_db = False):
+    def import_from_csv(self, state: ReplState, from_session: str, keyspace: str, table: str, target_table: str, columns: str, multi_tables = True, create_db = False):
         pass
 
-    def move_to_done(self, pod: str, namespace: str, to_session: str, from_session: str, keyspace: str, target_table: str):
+    @abstractmethod
+    def import_from_local_csv(self, state: ReplState,
+                        keyspace: str, table: str, target_table: str, columns: str,
+                        csv_file: str,
+                        multi_tables = True, create_db = False):
+        pass
+
+    def move_to_done(self, state: ReplState, from_session: str, keyspace: str, target_table: str):
+        pod = state.pod
+        namespace = state.namespace
+        to_session = state.export_session
         log_file = f'{log_prefix()}-{from_session}_{keyspace}.{target_table}.log.pending_import'
 
         to = f'{log_prefix()}-{to_session}_{keyspace}.{target_table}.log.done'
@@ -30,9 +41,12 @@ class Importer:
 
         return session
 
-    def remove_csv(self, pod: str, namespace: str, session: str, table: str, target_table: str, multi_tables = True):
-        with ing(f'[{session}] Cleaning up temporary files', suppress_log=multi_tables):
-            CassandraNodes.exec(pod, namespace, f'rm -rf {self.csv_file(session, table, target_table)}', show_out=Config().is_debug(), shell='bash')
+    def remove_csv(self, state: ReplState, from_session: str, table: str, target_table: str, multi_tables = True):
+        pod = state.pod
+        namespace = state.namespace
+
+        with ing(f'[{from_session}] Cleaning up temporary files', suppress_log=multi_tables):
+            CassandraNodes.exec(pod, namespace, f'rm -rf {self.csv_file(from_session, table, target_table)}', show_out=Config().is_debug(), shell='bash')
 
     def db(self, session: str, keyspace: str):
         return f'{session}_{keyspace}'