jbrowse2: jbrowse2.py comparison

comparison jbrowse2.py @ 80:dff27c9f6d72 draft

planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/jbrowse2 commit 7bb0fa234bdbc42768b30e426472a47b2523297f

author	fubar
date	Wed, 03 Apr 2024 01:00:21 +0000
parents	14ecbe46ae9f
children	e9bcbed05108

comparison

equal deleted inserted replaced

-:14ecbe46ae9f
+:dff27c9f6d72
 logging.basicConfig(level=logging.DEBUG)
 log = logging.getLogger("jbrowse")
 JB2VER = "v2.10.3"
-# version pinned for cloning
+# version pinned if cloning - but not cloning now
 TODAY = datetime.datetime.now().strftime("%Y-%m-%d")
 SELF_LOCATION = os.path.dirname(os.path.realpath(__file__))
 GALAXY_INFRASTRUCTURE_URL = None
 mapped_chars = {
 return metadata
 class JbrowseConnector(object):
 def __init__(self, outdir, jbrowse2path):
+self.trackCounter = 0  # to avoid name clashes
 self.assemblies = []  # these require more than a few line diff.
 self.assmeta = {}
 self.ass_first_contigs = (
 []
 )  # for default session - these are read as first line of the assembly .fai
 style_data,
 ]
 }
 return wstyle
-def urllib_get_2018():
+def getNrow(self, url):
-# Using a protected member like this is not any more fragile
+useuri = url.startswith("https://") or url.startswith("http://")
-# than extending the class and using it. I would use it.
+if not useuri:
-url = "https://localhost:6667/my-endpoint"
+fl = open(url, "r").readlines()
-ssl._create_default_https_context = ssl._create_unverified_context
+nrow = len(fl)
-with urllib.request.urlopen(url=url) as f:
+else:
-print(f.read().decode("utf-8"))
+try:
+scontext = ssl.SSLContext(ssl.PROTOCOL_TLS)
-def urllib_get_2022():
+scontext.verify_mode = ssl.VerifyMode.CERT_NONE
-# Finally! Able to use the publice API. Happy happy!
+with urllib.request.urlopen(url, context=scontext) as f:
-url = "https://localhost:6667/my-endpoint"
+fl = f.readlines()
-scontext = ssl.SSLContext(ssl.PROTOCOL_TLS)
+nrow = len(fl)
-scontext.verify_mode = ssl.VerifyMode.CERT_NONE
+except Exception:
-with urllib.request.urlopen(url=url, context=scontext) as f:
+nrow = 0
-print(f.read().decode("utf-8"))
+logging.debug("### getNrow returning %d" % nrow)
+return nrow
 def process_genomes(self, genomes):
 assembly = []
 assmeta = []
 useuri = False
 for i, genome_node in enumerate(genomes):
 this_genome = {}
 if genome_node["useuri"] == "yes":
 useuri = True
 genome_name = genome_node["label"].strip()
+if len(genome_name) == 0:
+genome_name = os.path.splitext(os.path.basename(genome_node["path"]))[0]
 if len(genome_name.split()) > 1:
 genome_name = genome_name.split()[0]
 # spaces and cruft break scripts when substituted
 if genome_name not in genome_names:
 # pafs with shared references
 # url = "%s/api/datasets/%s/display?to_ext=hic " % (self.giURL, dsId)
 useuri = trackData["useuri"].lower() == "yes"
 if useuri:
 uri = data
 else:
-uri = "%s.hic" % trackData["label"]
+uri = tId
 # slashes in names cause path trouble
 dest = os.path.join(self.outdir, uri)
 cmd = ["cp", data, dest]
 self.subprocess_check_call(cmd)
 categ = trackData["category"]
 "url": "https://unpkg.com/jbrowse-plugin-mafviewer/dist/jbrowse-plugin-mafviewer.umd.production.min.js",
 }
 ]
 }
 categ = trackData["category"]
-fname = "%s" % tId
+fname = tId
 dest = "%s/%s" % (self.outdir, fname)
 gname = trackData["assemblyNames"]
 cmd = [
 "bash",
 logging.debug("### gff3rebase cmd = %s" % " ".join(cmd))
 gff3_rebased.close()
 # Replace original gff3 file
 shutil.copy(gff3_rebased.name, gff3)
 os.unlink(gff3_rebased.name)
-url = "%s.gff3.gz" % trackData["label"]
+self.add_gff(gff3, trackData, **kwargs)
-dest = "%s/%s" % (self.outdir, url)
-self._sort_gff(gff3, dest)
+def add_bigwig(self, data, trackData):
 tId = trackData["label"]
-categ = trackData["category"]
-trackDict = {
-"type": "FeatureTrack",
-"trackId": tId,
-"name": trackData["name"],
-"assemblyNames": [trackData["assemblyNames"]],
-"category": [
-categ,
-],
-"adapter": {
-"type": "Gff3TabixAdapter",
-"gffGzLocation": {
-"uri": url,
-},
-"index": {
-"location": {
-"uri": url + ".tbi",
-}
-},
-},
-"displays": [
-{
-"type": "LinearBasicDisplay",
-"displayId": "%s-LinearBasicDisplay" % tId,
-},
-{
-"type": "LinearArcDisplay",
-"displayId": "%s-LinearArcDisplay" % tId,
-},
-],
-}
-style_json = self._prepare_track_style(trackDict)
-trackDict["style"] = style_json
-self.tracksToAdd[trackData["assemblyNames"]].append(trackDict)
-self.trackIdlist.append(tId)
-os.unlink(gff3)
-def add_bigwig(self, data, trackData):
 useuri = trackData["useuri"].lower() == "yes"
 if useuri:
 url = data
 else:
-url = "%s.bigwig" % trackData["label"]
+url = tId
 # slashes in names cause path trouble
 dest = os.path.join(self.outdir, url)
 cmd = ["cp", data, dest]
 self.subprocess_check_call(cmd)
 bwloc = {"uri": url}
-tId = trackData["label"]
 categ = trackData["category"]
 trackDict = {
 "type": "QuantitativeTrack",
 "trackId": tId,
 "name": trackData["name"],
 bindex = bam_index
 categ = trackData["category"]
 if useuri:
 url = data
 else:
-fname = "%s.bam" % trackData["label"]
+fname = tId
 dest = "%s/%s" % (self.outdir, fname)
 url = fname
 bindex = fname + ".bai"
 self.subprocess_check_call(["cp", data, dest])
 if bam_index is not None and os.path.exists(bam_index):
 genseqad = "Not found"
 logging.warn("No adapter found for cram %s in gsa=%s" % (tId, gsa))
 if useuri:
 url = data
 else:
-fname = "%s.cram" % trackData["label"]
+fname = tId
-dest = "%s/%s" % (self.outdir, fname)
+dest = os.path.join(self.outdir, fname)
 url = fname
 self.subprocess_check_call(["cp", data, dest])
 if cram_index is not None and os.path.exists(cram_index):
 if not os.path.exists(dest + ".crai"):
 # most probably made by galaxy and stored in galaxy dirs, need to copy it to dest
 categ = trackData["category"]
 useuri = trackData["useuri"].lower() == "yes"
 if useuri:
 url = data
 else:
-url = "%s.vcf.gz" % tId
+url = tId
 dest = "%s/%s" % (self.outdir, url)
 cmd = "bgzip -c %s  > %s" % (data, dest)
 self.subprocess_popen(cmd)
 cmd = ["tabix", "-f", "-p", "vcf", dest]
 self.subprocess_check_call(cmd)
 # Only index if not already done
 if not os.path.exists(dest):
 cmd = "jbrowse sort-gff '%s' | bgzip -c > '%s'" % (
 data,
 dest,
-)  # "gff3sort.pl --precise '%s' | grep -v \"^$\" > '%s'"
+)
 self.subprocess_popen(cmd)
 self.subprocess_check_call(["tabix", "-f", "-p", "gff", dest])
 def _sort_bed(self, data, dest):
 # Only index if not already done
 cmd = "sort -k1,1 -k2,2n '%s' | bgzip -c > '%s'" % (data, dest)
 self.subprocess_popen(cmd)
 cmd = ["tabix", "-f", "-p", "bed", dest]
 self.subprocess_check_call(cmd)
-def add_gff(self, data, ext, trackData):
+def add_gff(self, data, trackData):
+tId = trackData["label"]
 useuri = trackData["useuri"].lower() == "yes"
 if useuri:
 url = trackData["path"]
 else:
-url = "%s.%s.gz" % (trackData["label"], ext)
+url = tId + ".gz"
 dest = "%s/%s" % (self.outdir, url)
 self._sort_gff(data, dest)
-tId = trackData["label"]
 categ = trackData["category"]
 trackDict = {
 "type": "FeatureTrack",
 "trackId": tId,
 "name": trackData["name"],
 categ = trackData["category"]
 useuri = trackData["useuri"].lower() == "yes"
 if useuri:
 url = data
 else:
-url = "%s.%s.gz" % (trackData["label"], ext)
+url = tId
 dest = "%s/%s" % (self.outdir, url)
 self._sort_bed(data, dest)
 trackDict = {
 "type": "FeatureTrack",
 "trackId": tId,
 self.trackIdlist.append(tId)
 def add_paf(self, data, trackData, pafOpts, **kwargs):
 tname = trackData["name"]
 tId = trackData["label"]
-url = "%s.paf" % tId
+url = tId
 useuri = data.startswith("http://") or data.startswith("https://")
 if not useuri:
 dest = "%s/%s" % (self.outdir, url)
 self.symlink_or_copy(os.path.realpath(data), dest)
+nrow = self.getNrow(dest)
 else:
 url = data
+nrow = self.getNrow(url)
 categ = trackData["category"]
-pgnames = [
+pgnames = [x.strip() for x in pafOpts["genome_label"].split(",")]
-x.strip() for x in pafOpts["genome_label"].split(",") if len(x.strip()) > 0
-]
 pgpaths = [
 x.strip() for x in pafOpts["genome"].split(",") if len(x.strip()) > 0
 ]
 passnames = [trackData["assemblyNames"]]  # always first
+for i, gp in enumerate(pgpaths):
+if len(pgnames[i].strip()) == 0:
+# user may have left it blank - cannot make non-optional if want optional tracks.
+gn = os.path.basename(gp)
+pgnames[i] = os.path.splitext(gn)[0]
 logging.debug(
 "### add_paf got pafOpts=%s, pgnames=%s, pgpaths=%s for %s"
 % (pafOpts, pgnames, pgpaths, tId)
 )
-for i, gname in enumerate(pgnames):
+for i, gp in enumerate(pgpaths):
+gname = pgnames[i]
 if len(gname.split()) > 1:
 gname = gname.split()[0]
 passnames.append(gname)
 # trouble from spacey names in command lines avoidance
-useuri = pgpaths[i].startswith("http://") or pgpaths[i].startswith(
+useuri = gp.startswith("http://") or gp.startswith("https://")
-"https://"
-)
 if gname not in self.genome_names:
 # ignore if already there - eg for duplicates among pafs.
-asstrack, first_contig = self.make_assembly(pgpaths[i], gname, useuri)
+asstrack, first_contig = self.make_assembly(gp, gname, useuri)
 self.genome_names.append(gname)
 self.tracksToAdd[gname] = []
 self.assemblies.append(asstrack)
 trackDict = {
 "type": "SyntenyTrack",
 {
 "type": "LinearComparativeDisplay",
 "displayId": "%s-LinearComparativeDisplay" % tId,
 },
 {
-"type": "LinearSyntenyDisplay",
+"type": "LinearBasicDisplay",
 "displayId": "%s-LinearSyntenyDisplay" % tId,
 },
 ],
 }
-style_json = {
+if nrow > 50000:
-"displays": [
+style_json = {
-{
+"displays": [
-"type": "LGVSyntenyDisplay",
+{
-"displayId": "%s-LGVSyntenyDisplay" % tId,
+"type": "LGVSyntenyDisplay",
-}
+"displayId": "%s-LGVSyntenyDisplay" % tId,
-]
+}
-}
+]
+}
+else:
+style_json = {
+"displays": [
+{
+"type": "LinearBasicDisplay",
+"displayId": "%s-LinearBasicDisplay" % tId,
+}
+]
+}
 trackDict["style"] = style_json
 self.tracksToAdd[trackData["assemblyNames"]].append(trackDict)
 self.trackIdlist.append(tId)
 def process_annotations(self, track):
 category = track["category"].replace("__pd__date__pd__", TODAY)
-for i, (
+for trackIndex, (
 dataset_path,
 dataset_ext,
 useuri,
 track_human_label,
 extra_metadata,
 track_human_label = track_human_label.replace(" ", "_")
 outputTrackConfig = {
 "category": category,
 "style": {},
 }
 outputTrackConfig["assemblyNames"] = track["assemblyNames"]
 outputTrackConfig["key"] = track_human_label
 outputTrackConfig["useuri"] = useuri
 outputTrackConfig["path"] = dataset_path
 outputTrackConfig["ext"] = dataset_ext
 outputTrackConfig["trackset"] = track.get("trackset", {})
-outputTrackConfig["label"] = "%s_%i_%s" % (
+outputTrackConfig["label"] = "%s_%d.%s" % (
+track_human_label,
+self.trackCounter,
 dataset_ext,
-i,
-track_human_label,
 )
+self.trackCounter += 1
 outputTrackConfig["metadata"] = extra_metadata
 outputTrackConfig["name"] = track_human_label
 if dataset_ext in ("gff", "gff3"):
 self.add_gff(
 dataset_path,
-dataset_ext,
 outputTrackConfig,
 )
 elif dataset_ext in ("hic", "juicebox_hic"):
 self.add_hic(
 dataset_path,
 outputTrackConfig,
 )
 elif dataset_ext in ("cool", "mcool", "scool"):
-hic_url = "%s_%d.hic" % (track_human_label, i)
+hic_url = outputTrackConfig["label"]
 hic_path = os.path.join(self.outdir, hic_url)
 self.subprocess_check_call(
 [
 "hictk",
 "convert",
 logging.debug(
 "### No style data in default data %s for %s"
 % (default_data, tId)
 )
 style_data = {"type": "LinearBasicDisplay"}
 if "displays" in track_conf:
 disp = track_conf["displays"][0]["type"]
 style_data["type"] = disp
 if track_conf.get("style_labels", None):
 # TODO fix this: it should probably go in a renderer block (SvgFeatureRenderer) but still does not work
 # TODO move this to per track displays?
 style_data["labels"] = track_conf["style_labels"]
 tracks_data.append(
 first = [x for x in self.ass_first_contigs if x[0] == gnome]
 if len(first) > 0:
 [gnome, refName, end] = first[0]
 start = 0
 end = int(end)
-refName = self.assmeta[gnome][0].get("genome_firstcontig", None)
 drdict = {
 "refName": refName,
 "start": start,
 "end": end,
 "reversed": False,
 with open(self.config_json_file, "w") as config_file:
 json.dump(self.config_json, config_file, indent=2)
 def add_defsess_to_index(self, data):
 """
-This was included on request of the new codeowner from Anthony's IUC PR.
+Included on request of the new codeowner, from Anthony's IUC PR.
-Now fixed to deal with each assembly and tracks separately.
+Had to be fixed to keep each assembly with the associated tracks for a default view.
-Originally used only the first assembly, putting all tracks there and
+Originally used only the first assembly, putting all tracks there and so breaking some
-generally falling apart when tested with 2 or more. Seems ironic that
+when tested with 2 or more. Seems ironic that this vital feature could not have ever been tested
-this vital feature was never tested given the rejection of my original IUC PR
+given that my declining to add it was the basis for a reviewer's rejection of my original IUC PR.
-because it was not there. And no, reviewer, I do not want this important piece of history
+A simple 2 line diff apparently.
-removed. I prefer that it remain here since it has caused me considerable discomfort.
+The technical problem is that this index.html hack breaks the promise of all the form fields
+for track controls such as visibility default that were working mostly. They need to be removed from the form by whoever
+thought this method was a good solution to the JB2 bug breaking config.json style default
+view coordinates.
+And no, dear reviewer of this code, please leave this piece of history.
+It is true and I prefer that it remain here to document my considerable discomfort at this unfair treatment.
 ----------------------------------------------------------
 Add some default session settings: set some assemblies/tracks on/off
 This allows to select a default view:
 GALAXY_INFRASTRUCTURE_URL = "http://" + GALAXY_INFRASTRUCTURE_URL
 jc = JbrowseConnector(outdir=args.outdir, jbrowse2path=args.jbrowse2path)
 default_session_data = {}
 for ass in root.findall("assembly"):
 genomes = [
 {
 "path": x.attrib["path"],
 "label": x.attrib["label"],
 trackfiles = track.findall("files/trackFile")
 if trackfiles:
 for x in track.findall("files/trackFile"):
 track_conf["label"] = x.attrib["label"]
-trackkey = track_conf["label"]
 track_conf["useuri"] = x.attrib["useuri"]
 if is_multi_bigwig:
 multi_bigwig_paths.append(
 (
 x.attrib["label"],
 assconf = jc.config_json.get("assemblies", [])
 assconf += jc.assemblies
 jc.config_json["assemblies"] = assconf
 logging.debug("assemblies=%s, gnames=%s" % (assconf, jc.genome_names))
 jc.write_config()
-# jc.add_default_session(default_session_data)
+jc.add_default_session(default_session_data)
 # note that this can be left in the config.json but has NO EFFECT if add_defsess_to_index is called.
-jc.add_defsess_to_index(default_session_data)
+# jc.add_defsess_to_index(default_session_data)
 # jc.text_index() not sure what broke here.

Mercurial > repos > fubar > jbrowse2

comparison jbrowse2.py @ 80:dff27c9f6d72 draft