Commit f1580974 authored by Jan Kiene's avatar Jan Kiene
Browse files

Merge branch 'basop-ci/add-project-id-as-argument-for-get_id_of_last_job_occurence' into 'main' 

[BASOP-CI] basop-ci-related changes for test result presentation on gitlab pages

See merge request !1594
parents e662ac23 311fa583

.gitlab-ci.yml

+1 −1
Original line number Diff line number Diff line
@@ -2289,7 +2289,7 @@ coverage-test-on-main-scheduled: 
  &complexity-measurements-setup # create necessary environment

  - mkdir -p wmops/logs



  - job_id=$(python3 ci/get_id_of_last_job_occurence.py $CI_COMMIT_REF_NAME $CI_JOB_NAME)

  - job_id=$(python3 ci/get_id_of_last_job_occurence.py $CI_COMMIT_REF_NAME $CI_JOB_NAME $CI_PROJECT_ID --success_only)

  - echo $job_id

  - curl --request GET "https://forge.3gpp.org/rep/api/v4/projects/$CI_PROJECT_ID/jobs/$job_id/artifacts" --output artifacts.zip

  - unzip artifacts.zip || true # this may fail on first run, when there are no artifacts there and the zip file is actually just "404"-html

ci/basop-pages/basop_index.html

0 → 100644
+22 −0
Original line number Diff line number Diff line 
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">

<html>

  <head>

  </head>

<body>

  <h1>Ivas BASOP code Development</h1>



  <h2>Daily long testvector tests</h2>



  <ul>

    <li><a href="ivas-pytest-mld-long-dec-index.html">ivas-pytest-mld-long-dec</a></li>

    <li><a href="ivas-pytest-mld-long-dec-lev+10-index.html">ivas-pytest-mld-long-dec-lev+10</a></li>

    <li><a href="ivas-pytest-mld-long-dec-lev-10-index.html">ivas-pytest-mld-long-dec-lev-10</a></li>

  </ul>



  <h2>Test Coverage</h2>



  <br>

  tbd...

  </br>



</body>

ci/basop-pages/create_report_pages.py

0 → 100644
+219 −0
Original line number Diff line number Diff line 
import csv

import pathlib

import argparse





CSV_DELIM = ";"

SUBPAGE_TMPL_CSS = """

<style type="text/css">

.tbase  {border-collapse:collapse;border-spacing:0;}

.tbase td{border-color:black;border-style:solid;border-width:1px;font-family:sans-serif;font-size:14px;

  overflow:hidden;padding:10px 5px;word-break:normal;}

.tbase th{border-color:black;border-style:solid;border-width:1px;font-family:sans-serif;font-size:14px;

  font-weight:normal;overflow:hidden;padding:10px 5px;word-break:normal;}

.tbase .tunder{font-weight:bold;text-align:center;text-decoration:underline;vertical-align:top}

.tbase .tcenter{font-weight:bold;text-align:center;vertical-align:top}

.tbase .tleft{text-align:left;vertical-align:top}

.tbase .tincrease{text-align:left;vertical-align:top;background-color:#ff0000;border-color:inherit;font-weight:bold;}

.tbase .treduce{text-align:left;vertical-align:top;background-color:#00ff00;border-color:inherit;font-weight:bold;}

</style>

"""



SUBPAGE_TMPL_HTML = """



<h1>Report for job {job_name}</h1



Comparing:

<ul>

    <li>Current run - id: <a href="https://forge.3gpp.org/rep/sa4/audio/ivas-basop/-/jobs/{id_current}">{id_current}</a></li>

    <li>Previous run - id: <a href="https://forge.3gpp.org/rep/sa4/audio/ivas-basop/-/jobs/{id_previous}">{id_previous}</a></li>

    <li><a href="{job_name}--merged_csv--{id_current}--{id_previous}.csv">Merged csv data</a></li>

</ul>



<br>

<b>Table is sorted by Difference in MLD.</b>

<br>



<table class="tbase"><thead>

  <tr>

    <th class="tunder" rowspan="2">Testcase</th>

    <th class="tunder" colspan="2">MLD</th>

    <th class="tunder" colspan="2">Max Abs Diff</th>

  </tr>

  <tr>

    <th class="tcenter">{id_previous}</th>

    <th class="tcenter">{id_current}</th>

    <th class="tcenter">{id_previous}</th>

    <th class="tcenter">{id_current}</th>

  </tr></thead>

<tbody>

{table_body}

</tbody>

</table>

"""

TD_TMPL_NORMAL = "<td class='tleft'>{}</td>"

TD_TMPL_INCREASE = "<td class='tincrease'>{}</td>"

TD_TMPL_REDUCE = "<td class='treduce'>{}</td>"

TR_TMPL = "<tr>{}</tr>"



# expected columns. actual columns are filtered from the incoming data later, this

# is mainly for controlling the order in the output table

COLUMNS = ["testcase", "Result", "MLD", "MAXIMUM ABS DIFF"]

COLUMNS_GLOBAL = COLUMNS[:1]

COLUMNS_DIFFERENTIAL = COLUMNS[1:]





def create_subpage(

    html_out,

    csv_out,

    csv_current: str,

    csv_previous: str,

    id_current: int,

    id_previous: int,

    job_name: str,

):

    merged_reports = merge_and_cleanup_mld_reports(

        csv_current, csv_previous, id_current, id_previous

    )

    write_out_csv(merged_reports, merged_reports[0].keys(), csv_out)

    table_body = "\n".join(

        tr_from_row(row, id_current, id_previous) for row in merged_reports

    )

    new_subpage = SUBPAGE_TMPL_CSS + SUBPAGE_TMPL_HTML.format(

        id_current=id_current,

        id_previous=id_previous,

        table_body=table_body,

        job_name=job_name,

    )

    with open(html_out, "w") as f:

        f.write(new_subpage)





def write_out_csv(data, col_names, outfile):

    with open(outfile, "w") as f:

        writer = csv.DictWriter(f, col_names, delimiter=";")

        writer.writeheader()

        for row in data:

            writer.writerow(row)





def tr_from_row(row, id_current, id_previous):

    tr = list()



    # pre-filter columns to handle case where new columns are added

    # only include columns that are there in both data

    columns_global = [c for c in COLUMNS_GLOBAL if c in row]

    diff_col_tmpl = "{}-{}"

    incoming_cols = row.keys()

    columns_differential = [

        c

        for c in COLUMNS_DIFFERENTIAL

        if diff_col_tmpl.format(c, id_current) in incoming_cols

        and diff_col_tmpl.format(c, id_previous) in incoming_cols

    ]



    for c in columns_global:

        # this is currently for the "testcase" column - here we don't compare, just one value is used

        tr.append(TD_TMPL_NORMAL.format(row[c]))

    for c in columns_differential:

        # this is for all columns where we compare between current and previous run

        prev = row[f"{c}-{id_previous}"]

        curr = row[f"{c}-{id_current}"]



        # use red background if increase, green if decrease, white if same

        td_tmpl = TD_TMPL_NORMAL

        try:

            if float(curr) > float(prev):

                td_tmpl = TD_TMPL_INCREASE

            if float(curr) < float(prev):

                td_tmpl = TD_TMPL_REDUCE

        except ValueError:

            # if we land here, one of the cells is not a number, this indicates a crash

            # or some error in the scripts, so mark with red as well

            td_tmpl = TD_TMPL_INCREASE



        tr.append(td_tmpl.format(row[f"{c}-{id_previous}"]))

        tr.append(td_tmpl.format(row[f"{c}-{id_current}"]))



    return TR_TMPL.format("\n".join(tr))





def merge_and_cleanup_mld_reports(

    csv_current: str, csv_previous: str, id_current: int, id_previous: int

):

    with open(csv_current) as f:

        current_reader = csv.DictReader(f, delimiter=CSV_DELIM)

        current = list(current_reader)

    with open(csv_previous) as f:

        previous = list(csv.DictReader(f, delimiter=CSV_DELIM))



    # TODO: handle newly added testcases - for now assume that both have the same columns

    merge_key = "testcase"

    other_keys = [k for k in current_reader.fieldnames if k != merge_key]

    merged = merge_tables(

        current, previous, id_current, id_previous, merge_key, other_keys

    )



    # TODO: sort on result as well

    mld_col_curr = f"MLD-{id_current}"

    mld_col_prev = f"MLD-{id_previous}"



    # sort based on difference in MLD between current and previous run

    # put cases with "None" at the top of the list

    def sort_func(x):

        vals_missing = ["None", ""]



        if x[mld_col_curr] in vals_missing or x[mld_col_prev] in vals_missing:

            return float("inf")



        return float(x[mld_col_curr]) - float(x[mld_col_prev])



    merged = sorted(merged, key=sort_func, reverse=True)



    # remove the unecessary whole path from the testcase names

    for row in merged:

        row["testcase"] = pathlib.Path(row["testcase"]).name



    return merged





def merge_tables(tbl1, tbl2, suffix1, suffix2, merge_key, other_keys):

    merged = list()



    for row1 in tbl1:

        new_row = dict()

        for key in other_keys:

            new_row[f"{key}-{suffix1}"] = row1[key]



        for row2 in tbl2:

            if row1[merge_key] == row2[merge_key]:

                new_row[merge_key] = row1[merge_key]

                for key in other_keys:

                    new_row[f"{key}-{suffix2}"] = row2[key]

                break



        merged.append(new_row)



    return merged





if __name__ == "__main__":

    parser = argparse.ArgumentParser()

    parser.add_argument("html_out")

    parser.add_argument("csv_out")

    parser.add_argument("csv_current")

    parser.add_argument("csv_previous")

    parser.add_argument("id_current", type=int)

    parser.add_argument("id_previous", type=int)

    parser.add_argument("job_name")

    args = parser.parse_args()



    create_subpage(

        args.html_out,

        args.csv_out,

        args.csv_current,

        args.csv_previous,

        args.id_current,

        args.id_previous,

        args.job_name,

    )

ci/get_id_of_last_job_occurence.py

+38 −34
Original line number Diff line number Diff line
@@ -34,16 +34,17 @@ import argparse 


import requests





PER_PAGE_SUFFIX = "?per_page=50"

PAGE_SUFFIX = "&page={}"

API_BASE_URL = "https://forge.3gpp.org/rep/api/v4/projects/49"

API_URL_TMPL = "https://forge.3gpp.org/rep/api/v4/projects/{}/pipelines"





def get_job_id(branch_name, job_name):

def get_job_id(branch_name, job_name, project_id, success_only):

    job_id = -1

    # check last 500 pipelines max

    for page in range(100):

        url_pls = API_BASE_URL + "/pipelines"

        url_pls = API_URL_TMPL.format(project_id)



        # need both suffixes here to descend through the pages and get also older pipelines

        suffix = PER_PAGE_SUFFIX + PAGE_SUFFIX.format(page)
@@ -61,7 +62,8 @@ def get_job_id(branch_name, job_name): 


                # find actual job by name

                for job in resp_jobs.json():

                    if job["name"] == job_name and job["status"] == "success":

                    include_job = not success_only or job["status"] == "success"

                    if include_job and job["name"] == job_name:

                        job_id = job["id"]

                        break

                if job_id >= 0:
@@ -75,10 +77,12 @@ def get_job_id(branch_name, job_name): 


if __name__ == "__main__":

    parser = argparse.ArgumentParser()

    parser.add_argument("branch_name")

    parser.add_argument("job_name")

    parser.add_argument("branch_name", help="Name of the branch to search on")

    parser.add_argument("job_name", help="Name of the job to get the id of")

    parser.add_argument("project_id", help="ID of project to search in", type=int)

    parser.add_argument("--success_only", help="Only include jobs with status 'success'", action="store_true")



    args = parser.parse_args()



    job_id = get_job_id(args.branch_name, args.job_name)

    job_id = get_job_id(args.branch_name, args.job_name, args.project_id, args.success_only)

    print(job_id)

ci/setup_pages.py

+65 −21
Original line number Diff line number Diff line
@@ -3,10 +3,16 @@ import os 
import pathlib

import subprocess

import sys

import shutil

from tempfile import TemporaryDirectory



from get_id_of_last_job_occurence import get_job_id



JOBS = [

PROJECT_ID_FLOAT_REPO = 49

PROJECT_ID_BASOP_REPO = 77





JOBS_FLOAT_REPO = [

    "complexity-stereo-in-stereo-out",

    "complexity-ism-in-binaural-out",

    "complexity-sba-hoa3-in-hoa3-out",
@@ -15,44 +21,81 @@ JOBS = [ 
    "complexity-StereoDmxEVS-stereo-in-mono-out",

    "coverage-test-on-main-scheduled",

]

JOBS_BASOP_REPO = [

    "ivas-pytest-mld-long-dec",

]



JOBS_FOR_PROJECT_ID = {

    PROJECT_ID_FLOAT_REPO: JOBS_FLOAT_REPO,

    PROJECT_ID_BASOP_REPO: JOBS_BASOP_REPO,

}



ARTIFACTS = "artifacts.zip"

API_URL_BASE = "https://forge.3gpp.org/rep/api/v4/projects/{}/jobs"

PUBLIC = "./public"

PUBLIC_FOLDER = pathlib.Path("./public").absolute()





def main():

    PUBLIC_FOLDER.mkdir()



    public_folder = pathlib.Path(PUBLIC)

    public_folder.mkdir()

    project_id = int(os.environ["CI_PROJECT_ID"])

    jobs = JOBS_FOR_PROJECT_ID[project_id]

    success_only = project_id == PROJECT_ID_FLOAT_REPO

    failed_count = get_artifacts_for_jobs_and_return_num_failed(

        jobs, project_id, success_only

    )



    if failed_count == len(jobs):

        print("Artifact collection failed for all jobs to check.")

        sys.exit(1)



    index_html = PUBLIC_FOLDER.joinpath("index.html")

    if project_id == PROJECT_ID_FLOAT_REPO:

        src = pathlib.Path("ci/index-pages.html").absolute()

        shutil.move(src, index_html)

    elif project_id == PROJECT_ID_BASOP_REPO:

        src = pathlib.Path("ci/basop-pages/basop_index.html").absolute()

        shutil.move(src, index_html)



    sys.exit(0)





def get_artifacts_for_jobs_and_return_num_failed(

    jobs: list, project_id: int, success_only: bool

) -> int:

    """

    Get specified artifact folders for all jobs given and put them into the public folder.



    jobs: dictionary with the job names in the keys and a list of the

          public folders to copy from the artifacts in the values

          if "-public" is in the list, the actual folder name to copy is the key with "-public" appended

    """

    failed_count = 0

    for job in JOBS:

        job_id = get_job_id(os.environ["CI_COMMIT_REF_NAME"], job)



    for job in jobs:

        job_id = get_job_id( os.environ["CI_DEFAULT_BRANCH"], job, project_id, success_only)



        print(f"{job_id} - {job}")

        try:

            curl_for_artifacts(job_id)

            with TemporaryDirectory() as tmp_dir:

                curl_for_artifacts(job_id, tmp_dir)



            job_public = job + "-public"

            if job == "coverage-test-on-main-scheduled":

                job_public = "coverage"

                pathlib.Path("coverage_stv").rename(

                    public_folder.joinpath("coverage_stv")

                )

                tmp_dir = pathlib.Path(tmp_dir)



            pathlib.Path(job_public).rename(public_folder.joinpath(job_public))

                for artifact in tmp_dir.iterdir():

                    src = tmp_dir.joinpath(artifact).absolute()

                    dst = PUBLIC_FOLDER.joinpath(artifact.name)

                    print(f"{src} -> {dst}")

                    shutil.move(src, dst)



        except subprocess.CalledProcessError:

            print(f"Could not get artifacts for {job}")

            failed_count += 1



    if failed_count == len(JOBS):

        sys.exit(1)



    pathlib.Path("ci/index-pages.html").rename(public_folder.joinpath("index.html"))

    sys.exit(0)

    return failed_count





def curl_for_artifacts(job_id):

def curl_for_artifacts(job_id: int, exdir: str):

    cmd = [

        "curl",

        "--request",
@@ -61,6 +104,7 @@ def curl_for_artifacts(job_id): 
        "--output",

        ARTIFACTS,

    ]

    print(cmd)

    subprocess.run(cmd, check=True)



    # check for valid archive (if not, it is likely a 404 page, then display that)
@@ -73,7 +117,7 @@ def curl_for_artifacts(job_id): 
        raise subprocess.CalledProcessError(-1, "Unzip check failed")



    # do the actual unzipping

    cmd = ["unzip", ARTIFACTS]

    cmd = ["unzip", ARTIFACTS, "-d", exdir]

    subprocess.run(cmd, check=True)