recipetool: create: reimplement fetching with normal fetch/unpack tasks

Now that we have the ability to run the tasks in a more standard context
through tinfoil, change recipetool's fetching code to use that to fetch
files using it. This has the major advantage that any dependencies of
do_fetch and do_unpack (e.g. for subversion or npm) will be handled
automatically. This also has the beneficial side-effect of fixing a
recent regression that prevented this fetch operation from working with
memory resident bitbake.

Also fix devtool's usage of fetch_uri() at the same time so that we can
completely replace it.

Fixes [YOCTO #11710].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

4 files changed, 153 insertions, 65 deletions

diff --git a/scripts/lib/devtool/upgrade.py b/scripts/lib/devtool/upgrade.py
index 05fb9e5ed07..f077f37726e 100644
--- a/scripts/lib/devtool/upgrade.py
+++ b/scripts/lib/devtool/upgrade.py
@@ -207,10 +207,16 @@ def _extract_new_source(newpv, srctree, no_patch, srcrev, branch, keep_temp, tin
 
         tmpdir = tempfile.mkdtemp(prefix='devtool')
         try:
-            md5, sha256 = scriptutils.fetch_uri(tinfoil.config_data, uri, tmpdir, rev)
-        except bb.fetch2.FetchError as e:
+            checksums, ftmpdir = scriptutils.fetch_url(tinfoil, uri, rev, tmpdir, logger, preserve_tmp=keep_temp)
+        except scriptutils.FetchUrlFailure as e:
             raise DevtoolError(e)
 
+        if ftmpdir and keep_temp:
+            logger.info('Fetch temp directory is %s' % ftmpdir)
+
+        md5 = checksums['md5sum']
+        sha256 = checksums['sha256sum']
+
         tmpsrctree = _get_srctree(tmpdir)
         srctree = os.path.abspath(srctree)
 
diff --git a/scripts/lib/recipetool/create.py b/scripts/lib/recipetool/create.py
index 2b7cc76ccbd..2a6a28ba91d 100644
--- a/scripts/lib/recipetool/create.py
+++ b/scripts/lib/recipetool/create.py
@@ -417,7 +417,7 @@ def create_recipe(args):
         pkgarch = "${MACHINE_ARCH}"
 
     extravalues = {}
-    checksums = (None, None)
+    checksums = {}
     tempsrc = ''
     source = args.source
     srcsubdir = ''
@@ -439,22 +439,25 @@ def create_recipe(args):
         if res:
             srcrev = res.group(1)
             srcuri = rev_re.sub('', srcuri)
-        tempsrc = tempfile.mkdtemp(prefix='recipetool-')
-        srctree = tempsrc
-        d = bb.data.createCopy(tinfoil.config_data)
-        if fetchuri.startswith('npm://'):
-            # Check if npm is available
-            npm_bindir = check_npm(tinfoil, args.devtool)
-            d.prependVar('PATH', '%s:' % npm_bindir)
-        logger.info('Fetching %s...' % srcuri)
+
+        tmpparent = tinfoil.config_data.getVar('BASE_WORKDIR')
+        bb.utils.mkdirhier(tmpparent)
+        tempsrc = tempfile.mkdtemp(prefix='recipetool-', dir=tmpparent)
+        srctree = os.path.join(tempsrc, 'source')
+
         try:
-            checksums = scriptutils.fetch_uri(d, fetchuri, srctree, srcrev)
-        except bb.fetch2.BBFetchException as e:
-            logger.error(str(e).rstrip())
+            checksums, ftmpdir = scriptutils.fetch_url(tinfoil, srcuri, srcrev, srctree, logger, preserve_tmp=args.keep_temp)
+        except scriptutils.FetchUrlFailure as e:
+            logger.error(str(e))
             sys.exit(1)
+
+        if ftmpdir and args.keep_temp:
+            logger.info('Fetch temp directory is %s' % ftmpdir)
+
         dirlist = os.listdir(srctree)
-        if 'git.indirectionsymlink' in dirlist:
-            dirlist.remove('git.indirectionsymlink')
+        filterout = ['git.indirectionsymlink']
+        dirlist = [x for x in dirlist if x not in filterout]
+        logger.debug('Directory listing (excluding filtered out):\n  %s' % '\n  '.join(dirlist))
         if len(dirlist) == 1:
             singleitem = os.path.join(srctree, dirlist[0])
             if os.path.isdir(singleitem):
@@ -465,7 +468,7 @@ def create_recipe(args):
                 check_single_file(dirlist[0], fetchuri)
         elif len(dirlist) == 0:
             if '/' in fetchuri:
-                fn = os.path.join(d.getVar('DL_DIR'), fetchuri.split('/')[-1])
+                fn = os.path.join(tinfoil.config_data.getVar('DL_DIR'), fetchuri.split('/')[-1])
                 if os.path.isfile(fn):
                     check_single_file(fn, fetchuri)
             # If we've got to here then there's no source so we might as well give up
@@ -593,11 +596,8 @@ def create_recipe(args):
     if not srcuri:
         lines_before.append('# No information for SRC_URI yet (only an external source tree was specified)')
     lines_before.append('SRC_URI = "%s"' % srcuri)
-    (md5value, sha256value) = checksums
-    if md5value:
-        lines_before.append('SRC_URI[md5sum] = "%s"' % md5value)
-    if sha256value:
-        lines_before.append('SRC_URI[sha256sum] = "%s"' % sha256value)
+    for key, value in sorted(checksums.items()):
+        lines_before.append('SRC_URI[%s] = "%s"' % (key, value))
     if srcuri and supports_srcrev(srcuri):
         lines_before.append('')
         lines_before.append('# Modify these as desired')
diff --git a/scripts/lib/recipetool/create_npm.py b/scripts/lib/recipetool/create_npm.py
index cb8f338b8b3..ba7e39a4067 100644
--- a/scripts/lib/recipetool/create_npm.py
+++ b/scripts/lib/recipetool/create_npm.py
@@ -109,7 +109,6 @@ class NpmRecipeHandler(RecipeHandler):
             if varname == 'SRC_URI':
                 if not origvalue.startswith('npm://'):
                     src_uri = origvalue.split()
-                    changed = False
                     deplist = {}
                     for dep, depver in optdeps.items():
                         depdata = self.get_npm_data(dep, depver, d)
@@ -123,14 +122,15 @@ class NpmRecipeHandler(RecipeHandler):
                         depdata = self.get_npm_data(dep, depver, d)
                         deplist[dep] = depdata
 
+                    extra_urls = []
                     for dep, depdata in deplist.items():
                         version = depdata.get('version', None)
                         if version:
                             url = 'npm://registry.npmjs.org;name=%s;version=%s;subdir=node_modules/%s' % (dep, version, dep)
-                            scriptutils.fetch_uri(d, url, srctree)
-                            src_uri.append(url)
-                            changed = True
-                    if changed:
+                            extra_urls.append(url)
+                    if extra_urls:
+                        scriptutils.fetch_url(tinfoil, ' '.join(extra_urls), None, srctree, logger)
+                        src_uri.extend(extra_urls)
                         return src_uri, None, -1, True
             return origvalue, None, 0, True
         updated, newlines = bb.utils.edit_metadata(lines_before, ['SRC_URI'], varfunc)
diff --git a/scripts/lib/scriptutils.py b/scripts/lib/scriptutils.py
index 92b601c7e84..1005dd495ad 100644
--- a/scripts/lib/scriptutils.py
+++ b/scripts/lib/scriptutils.py
@@ -23,6 +23,8 @@ import argparse
 import subprocess
 import tempfile
 import shutil
+import random
+import string
 
 def logger_create(name, stream=None):
     logger = logging.getLogger(name)
@@ -78,50 +80,130 @@ def git_convert_standalone_clone(repodir):
             bb.process.run('git repack -a', cwd=repodir)
             os.remove(alternatesfile)
 
-def fetch_uri(d, uri, destdir, srcrev=None):
-    """Fetch a URI to a local directory"""
+def _get_temp_recipe_dir(d):
+    # This is a little bit hacky but we need to find a place where we can put
+    # the recipe so that bitbake can find it. We're going to delete it at the
+    # end so it doesn't really matter where we put it.
+    bbfiles = d.getVar('BBFILES').split()
+    fetchrecipedir = None
+    for pth in bbfiles:
+        if pth.endswith('.bb'):
+            pthdir = os.path.dirname(pth)
+            if os.access(os.path.dirname(os.path.dirname(pthdir)), os.W_OK):
+                fetchrecipedir = pthdir.replace('*', 'recipetool')
+                if pthdir.endswith('workspace/recipes/*'):
+                    # Prefer the workspace
+                    break
+    return fetchrecipedir
+
+class FetchUrlFailure(Exception):
+    def __init__(self, url):
+        self.url = url
+    def __str__(self):
+        return "Failed to fetch URL %s" % self.url
+
+def fetch_url(tinfoil, srcuri, srcrev, destdir, logger, preserve_tmp=False):
+    """
+    Fetch the specified URL using normal do_fetch and do_unpack tasks, i.e.
+    any dependencies that need to be satisfied in order to support the fetch
+    operation will be taken care of
+    """
+
     import bb
-    tmpparent = d.getVar('BASE_WORKDIR')
+
+    checksums = {}
+    fetchrecipepn = None
+
+    # We need to put our temp directory under ${BASE_WORKDIR} otherwise
+    # we may have problems with the recipe-specific sysroot population
+    tmpparent = tinfoil.config_data.getVar('BASE_WORKDIR')
     bb.utils.mkdirhier(tmpparent)
-    tmpworkdir = tempfile.mkdtemp(dir=tmpparent)
+    tmpdir = tempfile.mkdtemp(prefix='recipetool-', dir=tmpparent)
     try:
-        bb.utils.mkdirhier(destdir)
-        localdata = bb.data.createCopy(d)
-
-        # Set some values to allow extend_recipe_sysroot to work here we're we are not running from a task
-        localdata.setVar('WORKDIR', tmpworkdir)
-        localdata.setVar('BB_RUNTASK', 'do_fetch')
-        localdata.setVar('PN', 'dummy')
-        localdata.setVar('BB_LIMITEDDEPS', '1')
-        bb.build.exec_func("extend_recipe_sysroot", localdata)
-
-        # Set some values for the benefit of the fetcher code
-        localdata.setVar('BB_STRICT_CHECKSUM', '')
-        localdata.setVar('SRCREV', srcrev)
-        ret = (None, None)
-        olddir = os.getcwd()
+        tmpworkdir = os.path.join(tmpdir, 'work')
+        logger.debug('fetch_url: temp dir is %s' % tmpdir)
+
+        fetchrecipedir = _get_temp_recipe_dir(tinfoil.config_data)
+        if not fetchrecipedir:
+            logger.error('Searched BBFILES but unable to find a writeable place to put temporary recipe')
+            sys.exit(1)
+        fetchrecipe = None
+        bb.utils.mkdirhier(fetchrecipedir)
         try:
-            fetcher = bb.fetch2.Fetch([uri], localdata)
-            for u in fetcher.ud:
-                ud = fetcher.ud[u]
-                ud.ignore_checksums = True
-            fetcher.download()
-            for u in fetcher.ud:
-                ud = fetcher.ud[u]
-                if ud.localpath.rstrip(os.sep) == localdata.getVar('DL_DIR').rstrip(os.sep):
-                    raise Exception('Local path is download directory - please check that the URI "%s" is correct' % uri)
-            fetcher.unpack(destdir)
-            for u in fetcher.ud:
-                ud = fetcher.ud[u]
-                if ud.method.recommends_checksum(ud):
-                    md5value = bb.utils.md5_file(ud.localpath)
-                    sha256value = bb.utils.sha256_file(ud.localpath)
-                    ret = (md5value, sha256value)
+            # Generate a dummy recipe so we can follow more or less normal paths
+            # for do_fetch and do_unpack
+            # I'd use tempfile functions here but underscores can be produced by that and those
+            # aren't allowed in recipe file names except to separate the version
+            rndstring = ''.join(random.choice(string.ascii_uppercase + string.digits) for _ in range(8))
+            fetchrecipe = os.path.join(fetchrecipedir, 'tmp-recipetool-%s.bb' % rndstring)
+            fetchrecipepn = os.path.splitext(os.path.basename(fetchrecipe))[0]
+            logger.debug('Generating initial recipe %s for fetching' % fetchrecipe)
+            with open(fetchrecipe, 'w') as f:
+                # We don't want to have to specify LIC_FILES_CHKSUM
+                f.write('LICENSE = "CLOSED"\n')
+                # We don't need the cross-compiler
+                f.write('INHIBIT_DEFAULT_DEPS = "1"\n')
+                # We don't have the checksums yet so we can't require them
+                f.write('BB_STRICT_CHECKSUM = "ignore"\n')
+                f.write('SRC_URI = "%s"\n' % srcuri)
+                f.write('SRCREV = "%s"\n' % srcrev)
+                f.write('WORKDIR = "%s"\n' % tmpworkdir)
+                # Set S out of the way so it doesn't get created under the workdir
+                f.write('S = "%s"\n' % os.path.join(tmpdir, 'emptysrc'))
+
+            logger.info('Fetching %s...' % srcuri)
+
+            # FIXME this is too noisy at the moment
+
+            # Parse recipes so our new recipe gets picked up
+            tinfoil.parse_recipes()
+
+            def eventhandler(event):
+                if isinstance(event, bb.fetch2.MissingChecksumEvent):
+                    checksums.update(event.checksums)
+                    return True
+                return False
+
+            # Run the fetch + unpack tasks
+            res = tinfoil.build_targets(fetchrecipepn,
+                                        'do_unpack',
+                                        handle_events=True,
+                                        extra_events=['bb.fetch2.MissingChecksumEvent'],
+                                        event_callback=eventhandler)
+            if not res:
+                raise FetchUrlFailure(srcuri)
+
+            # Remove unneeded directories
+            rd = tinfoil.parse_recipe(fetchrecipepn)
+            if rd:
+                pathvars = ['T', 'RECIPE_SYSROOT', 'RECIPE_SYSROOT_NATIVE']
+                for pathvar in pathvars:
+                    path = rd.getVar(pathvar)
+                    shutil.rmtree(path)
         finally:
-            os.chdir(olddir)
+            if fetchrecipe:
+                try:
+                    os.remove(fetchrecipe)
+                except FileNotFoundError:
+                    pass
+            try:
+                os.rmdir(fetchrecipedir)
+            except OSError as e:
+                import errno
+                if e.errno != errno.ENOTEMPTY:
+                    raise
+
+        bb.utils.mkdirhier(destdir)
+        for fn in os.listdir(tmpworkdir):
+            shutil.move(os.path.join(tmpworkdir, fn), destdir)
+
     finally:
-        shutil.rmtree(tmpworkdir)
-    return ret
+        if not preserve_tmp:
+            shutil.rmtree(tmpdir)
+            tmpdir = None
+
+    return checksums, tmpdir
+
 
 def run_editor(fn):
     if isinstance(fn, str):