Some simplifications to Scrapyd architecture and internals:

- launcher no longer knows about egg storage
- removed get_spider_list_from_eggifile() file and replaced by simpler
  get_spider_list() which doesn't receive en egg file as argument
- changed "egg runner" name to just "runner" to reflect the fact that it
  doesn't necesarilly run eggs (though it does in the default case)

--HG--
rename : scrapyd/eggrunner.py => scrapyd/runner.py

scrapyd/default_scrapyd.conf

@@ -7,5 +7,5 @@ max_proc    = 0
 max_proc_per_cpu = 4
 http_port   = 6800
 debug       = off
-egg_runner  = scrapyd.eggrunner
+runner      = scrapyd.runner
 application = scrapyd.app.application

scrapyd/eggutils.py

@@ -1,24 +1,4 @@
-from __future__ import with_statement
-
-import os, sys, shutil, pkg_resources
-from subprocess import Popen, PIPE
-from tempfile import NamedTemporaryFile
-
-def get_spider_list_from_eggfile(eggfile, project, eggrunner='scrapyd.eggrunner'):
-    with NamedTemporaryFile(suffix='.egg') as f:
-        shutil.copyfileobj(eggfile, f)
-        f.flush()
-        eggfile.seek(0)
-        pargs = [sys.executable, '-m', eggrunner, 'list']
-        env = os.environ.copy()
-        env['SCRAPY_PROJECT'] = project
-        env['SCRAPY_EGGFILE'] = f.name
-        proc = Popen(pargs, stdout=PIPE, stderr=PIPE, env=env)
-        out, err = proc.communicate()
-        if proc.returncode:
-            msg = err or out or 'unknown error'
-            raise RuntimeError(msg.splitlines()[-1])
-        return out.splitlines()
+import os, pkg_resources
 
 def activate_egg(eggpath):
     """Activate a Scrapy egg file. This is meant to be used from egg runners

scrapyd/environ.py

@@ -18,16 +18,14 @@ def __init__(self, config, initenv=os.environ):
             self.settings = {}
         self.initenv = initenv
 
-    def get_environment(self, message, slot, eggpath):
+    def get_environment(self, message, slot):
         project = message['_project']
         env = self.initenv.copy()
         env['SCRAPY_SLOT'] = str(slot)
         env['SCRAPY_PROJECT'] = project
         env['SCRAPY_SPIDER'] = message['_spider']
         env['SCRAPY_JOB'] = message['_job']
-        if eggpath:
-            env['SCRAPY_EGGFILE'] = eggpath
-        elif project in self.settings:
+        if project in self.settings:
             env['SCRAPY_SETTINGS_MODULE'] = self.settings[project]
         dbpath = os.path.join(self.dbs_dir, '%s.db' % project)
         env['SCRAPY_SQLITE_DB'] = dbpath

scrapyd/interfaces.py

@@ -64,13 +64,9 @@ def update_projects():
 class IEnvironment(Interface):
     """A component to generate the environment of crawler processes"""
 
-    def get_environment(message, slot, eggpath):
+    def get_environment(message, slot):
         """Return the environment variables to use for running the process.
 
         `message` is the message received from the IPoller.next() method
         `slot` is the Launcher slot where the process will be running.
-        `eggpath` is the path to an eggfile that contains the project code. The
-           `eggpath` may be `None` if no egg was found for the project, in
-           which case the project must be on the python path and its settings
-           defined in scrapyd.conf [settings] section
         """

scrapyd/launcher.py

@@ -1,6 +1,4 @@
 import sys, os
-from shutil import copyfileobj
-from tempfile import mkstemp
 from datetime import datetime
 
 from twisted.internet import reactor, defer, protocol, error
@@ -10,7 +8,7 @@
 from scrapy.utils.py26 import cpu_count
 from scrapy.utils.python import stringify_dict
 from scrapyd.utils import get_crawl_args
-from .interfaces import IPoller, IEggStorage, IEnvironment
+from .interfaces import IPoller, IEnvironment
 
 class Launcher(Service):
 
@@ -21,57 +19,41 @@ def __init__(self, config, app):
         self.max_proc = config.getint('max_proc', 0)
         if not self.max_proc:
             self.max_proc = cpu_count() * config.getint('max_proc_per_cpu', 4)
-        self.egg_runner = config.get('egg_runner', 'scrapyd.eggrunner')
+        self.runner = config.get('runner', 'scrapyd.runner')
         self.app = app
 
     def startService(self):
         for slot in range(self.max_proc):
             self._wait_for_project(slot)
-        log.msg("%s started: max_proc=%r, egg_runner=%r" % (self.parent.name, \
-            self.max_proc, self.egg_runner), system="Launcher")
+        log.msg("%s started: max_proc=%r, runner=%r" % (self.parent.name, \
+            self.max_proc, self.runner), system="Launcher")
 
     def _wait_for_project(self, slot):
         poller = self.app.getComponent(IPoller)
         poller.next().addCallback(self._spawn_process, slot)
 
-    def _get_eggpath(self, project):
-        eggstorage = self.app.getComponent(IEggStorage)
-        version, eggf = eggstorage.get(project)
-        if eggf is None:
-            return
-        prefix = '%s-%s-' % (project, version)
-        fd, eggpath = mkstemp(prefix=prefix, suffix='.egg')
-        lf = os.fdopen(fd, 'wb')
-        copyfileobj(eggf, lf)
-        lf.close()
-        return eggpath
-
     def _spawn_process(self, message, slot):
         msg = stringify_dict(message, keys_only=False)
         project = msg['_project']
-        eggpath = self._get_eggpath(project)
-        args = [sys.executable, '-m', self.egg_runner, 'crawl']
+        args = [sys.executable, '-m', self.runner, 'crawl']
         args += get_crawl_args(msg)
         e = self.app.getComponent(IEnvironment)
-        env = e.get_environment(msg, slot, eggpath)
+        env = e.get_environment(msg, slot)
         env = stringify_dict(env, keys_only=False)
-        pp = ScrapyProcessProtocol(eggpath, slot, project, msg['_spider'], \
+        pp = ScrapyProcessProtocol(slot, project, msg['_spider'], \
             msg['_job'], env)
-        pp.deferred.addBoth(self._process_finished, eggpath, slot)
+        pp.deferred.addBoth(self._process_finished, slot)
         reactor.spawnProcess(pp, sys.executable, args=args, env=env)
         self.processes[slot] = pp
 
-    def _process_finished(self, _, eggpath, slot):
-        if eggpath:
-            os.remove(eggpath)
+    def _process_finished(self, _, slot):
         self.processes.pop(slot)
         self._wait_for_project(slot)
 
 
 class ScrapyProcessProtocol(protocol.ProcessProtocol):
 
-    def __init__(self, eggfile, slot, project, spider, job, env):
-        self.eggfile = eggfile
+    def __init__(self, slot, project, spider, job, env):
         self.slot = slot
         self.pid = None
         self.project = project
@@ -100,6 +82,6 @@ def processEnded(self, status):
         self.deferred.callback(self)
 
     def log(self, msg):
-        msg += "project=%r spider=%r job=%r pid=%r egg=%r log=%r" % (self.project, \
-            self.spider, self.job, self.pid, self.eggfile, self.logfile)
+        msg += "project=%r spider=%r job=%r pid=%r log=%r" % (self.project, \
+            self.spider, self.job, self.pid, self.logfile)
         log.msg(msg, system="Launcher")

scrapyd/runner.py

@@ -0,0 +1,37 @@
+import os
+import shutil
+import tempfile
+from contextlib import contextmanager
+
+from scrapyd import get_application
+from scrapyd.interfaces import IEggStorage
+from scrapyd.eggutils import activate_egg
+
+@contextmanager
+def project_environment(project):
+    app = get_application()
+    eggstorage = app.getComponent(IEggStorage)
+    version, eggfile = eggstorage.get(project)
+    if eggfile:
+        prefix = '%s-%s-' % (project, version)
+        fd, eggpath = tempfile.mkstemp(prefix=prefix, suffix='.egg')
+        lf = os.fdopen(fd, 'wb')
+        shutil.copyfileobj(eggfile, lf)
+        lf.close()
+        activate_egg(eggpath)
+    else:
+        eggpath = None
+    try:
+        yield
+    finally:
+        if eggpath:
+            os.remove(eggpath)
+
+def main():
+    project = os.environ['SCRAPY_PROJECT']
+    with project_environment(project):
+        from scrapy.cmdline import execute
+        execute()
+
+if __name__ == '__main__':
+    main()

scrapyd/tests/test_environ.py

@@ -8,7 +8,7 @@
 from scrapyd.config import Config
 from scrapyd.environ import Environment
 
-class EggStorageTest(unittest.TestCase):
+class EnvironmentTest(unittest.TestCase):
 
     def setUp(self):
         d = self.mktemp()
@@ -24,25 +24,11 @@ def test_interface(self):
     def test_get_environment_with_eggfile(self):
         msg = {'_project': 'mybot', '_spider': 'myspider', '_job': 'ID'}
         slot = 3
-        env = self.environ.get_environment(msg, slot, '/path/to/file.egg')
+        env = self.environ.get_environment(msg, slot)
         self.assertEqual(env['SCRAPY_PROJECT'], 'mybot')
         self.assertEqual(env['SCRAPY_SLOT'], '3')
         self.assertEqual(env['SCRAPY_SPIDER'], 'myspider')
         self.assertEqual(env['SCRAPY_JOB'], 'ID')
         self.assert_(env['SCRAPY_SQLITE_DB'].endswith('mybot.db'))
         self.assert_(env['SCRAPY_LOG_FILE'].endswith('/mybot/myspider/ID.log'))
-        self.assert_(env['SCRAPY_EGGFILE'].endswith('/path/to/file.egg'))
         self.failIf('SCRAPY_SETTINGS_MODULE' in env)
-
-    def test_get_environment_without_eggfile(self):
-        msg = {'_project': 'newbot', '_spider': 'myspider', '_job': 'ID'}
-        slot = 3
-        env = self.environ.get_environment(msg, slot, None)
-        self.assertEqual(env['SCRAPY_PROJECT'], 'newbot')
-        self.assertEqual(env['SCRAPY_SLOT'], '3')
-        self.assertEqual(env['SCRAPY_SPIDER'], 'myspider')
-        self.assertEqual(env['SCRAPY_JOB'], 'ID')
-        self.assert_(env['SCRAPY_SQLITE_DB'].endswith('newbot.db'))
-        self.assert_(env['SCRAPY_LOG_FILE'].endswith('/newbot/myspider/ID.log'))
-        self.assertEqual(env['SCRAPY_SETTINGS_MODULE'], 'newbot.settings')
-        self.failIf('SCRAPY_EGGFILE' in env)

scrapyd/tests/test_utils.py

@@ -1,6 +1,16 @@
-import unittest
+from __future__ import with_statement
 
-from scrapyd.utils import get_crawl_args
+import os
+from cStringIO import StringIO
+
+from twisted.trial import unittest
+
+from scrapy.utils.py26 import get_data
+from scrapyd.interfaces import IEggStorage
+from scrapyd.utils import get_crawl_args, get_spider_list
+from scrapyd import get_application
+
+__package__ = 'scrapyd.tests' # required for compatibility with python 2.5
 
 class UtilsTest(unittest.TestCase):
 
@@ -11,3 +21,27 @@ def test_get_crawl_args(self):
         cargs = get_crawl_args(msg)
         self.assertEqual(cargs, ['lala', '-a', 'arg1=val1'])
         assert all(isinstance(x, str) for x in cargs), cargs
+
+class GetSpiderListTest(unittest.TestCase):
+
+    def test_get_spider_list(self):
+        path = self.mktemp()
+        j = os.path.join
+        eggs_dir = j(path, 'eggs')
+        os.makedirs(eggs_dir)
+        dbs_dir = j(path, 'dbs')
+        os.makedirs(dbs_dir)
+        logs_dir = j(path, 'logs')
+        os.makedirs(logs_dir)
+        os.chdir(path)
+        with open('scrapyd.conf', 'w') as f:
+            f.write("[scrapyd]\n")
+            f.write("eggs_dir = %s\n" % eggs_dir)
+            f.write("dbs_dir = %s\n" % dbs_dir)
+            f.write("logs_dir = %s\n" % logs_dir)
+        app = get_application()
+        eggstorage = app.getComponent(IEggStorage)
+        eggfile = StringIO(get_data(__package__, 'mybot.egg'))
+        eggstorage.put(eggfile, 'mybot', 'r1')
+        self.assertEqual(sorted(get_spider_list('mybot')), ['spider1', 'spider2'])
+

scrapyd/utils.py

@@ -1,4 +1,6 @@
+import sys
 import os
+from subprocess import Popen, PIPE
 from ConfigParser import NoSectionError
 
 from scrapy.spiderqueue import SqliteSpiderQueue
@@ -41,3 +43,16 @@ def get_crawl_args(message):
         args += ['-a']
         args += ['%s=%s' % (k, v)]
     return args
+
+def get_spider_list(project, runner='scrapyd.runner'):
+    """Return the spider list from the given project, using the given runner"""
+    env = os.environ.copy()
+    env['SCRAPY_PROJECT'] = project
+    pargs = [sys.executable, '-m', runner, 'list']
+    proc = Popen(pargs, stdout=PIPE, stderr=PIPE, env=env)
+    out, err = proc.communicate()
+    if proc.returncode:
+        msg = err or out or 'unknown error'
+        raise RuntimeError(msg.splitlines()[-1])
+    return out.splitlines()
+

scrapyd/webservice.py

@@ -4,7 +4,7 @@
 from cStringIO import StringIO
 
 from scrapy.utils.txweb import JsonResource
-from .eggutils import get_spider_list_from_eggfile
+from .utils import get_spider_list
 
 class WsResource(JsonResource):
 
@@ -42,8 +42,8 @@ def render_POST(self, txrequest):
         project = d['project'][0]
         version = d['version'][0]
         eggf = StringIO(d['egg'][0])
-        spiders = get_spider_list_from_eggfile(eggf, project)
         self.root.eggstorage.put(eggf, project, version)
+        spiders = get_spider_list(project)
         self.root.update_projects()
         return {"status": "ok", "project": project, "version": version, \
             "spiders": len(spiders)}
@@ -65,9 +65,7 @@ class ListSpiders(WsResource):
 
     def render_GET(self, txrequest):
         project = txrequest.args['project'][0]
-        _, eggf = self.root.eggstorage.get(project)
-        spiders = get_spider_list_from_eggfile(eggf, project, \
-            eggrunner=self.root.egg_runner)
+        spiders = get_spider_list(project, runner=self.root.runner)
         return {"status": "ok", "spiders": spiders}
 
 class DeleteProject(WsResource):

scrapyd/website.py

@@ -11,7 +11,7 @@ class Root(resource.Resource):
     def __init__(self, config, app):
         resource.Resource.__init__(self)
         self.debug = config.getboolean('debug', False)
-        self.eggrunner = config.get('egg_runner')
+        self.runner = config.get('runner')
         logsdir = config.get('logs_dir')
         self.app = app
         self.putChild('', Home(self))