Split metadata template

[josenavas]

This is the first of multiple PR that are going to implement all the functionality included in the giant PR #1021.

I split the metadata_template.py file in 5 files (similarly for the test code):

• metadata_template/base_metadata_template.py: base classes
• metadata_template/sample_template.py: Sample and SampleTemplate classes
• metadata_template/prep_template.py: PrepSample and PrepTemplate classes
• metadata_template/util.py: All util functions
• metadata_template/constants.py: Hold constant values. This will be useful in the refactor, although it looks silly now.

I wanted to modify the minimal amount of code but I had to made a couple of changes:

• Some functions in metadata_template/util.py had a leading underscores on their name because they're private to the module. Since now they live in their own module, I remove that leading underscore (just a name change)
• The function create_qiime_mapping_file in the PrepTemplate class was using the SampleTemplate class. I changed this lines:

# getting the latest sample template
_, sample_template_fp = SampleTemplate(
    self.study_id).get_filepaths()[0]

to

# getting the latest sample template
conn_handler = SQLConnectionHandler()
sql = """SELECT filepath_id, filepath
         FROM qiita.filepath
            JOIN qiita.sample_template_filepath
            USING (filepath_id)
         WHERE study_id=%s
         ORDER BY filepath_id DESC"""
sample_template_fname = conn_handler.execute_fetchall(
    sql, (self.study_id,))[0][1]
_, fp = get_mountpoint('templates')[0]
sample_template_fp = join(fp, sample_template_fname)

to avoid the circular import.

More changes are required in the metadata template, I will start submitting PR based on this one starting tomorrow, so having this merged is important.

[coveralls]

Coverage increased (+0.11%) to 79.23% when pulling cd35e3f on josenavas:split-metadata-template into 9350269 on biocore:master.

[antgonza]

👍

[ElDeveloper]

🎆

On (Apr-02-15| 8:18), Joshua Shorenstein wrote:

Merged #1030.

---

Reply to this email directly or view it on GitHub:
#1030 (comment)