Skip to content

fill_metadata

ensembl.brc4.runnable.fill_metadata

fill_metadata

Bases: BaseRunnable

Source code in src/python/ensembl/brc4/runnable/fill_metadata.py
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
class fill_metadata(eHive.BaseRunnable):
    def param_defaults(self):
        return {
            "division": "",
            "ignore": [],
            "copy": {},
        }

    def run(self):
        # params
        wd = self.param_required("work_dir")
        division = self.param("division")
        brc4_mode = self.param("brc4_mode")
        ignore = self.param("ignore")
        copy = self.param("copy")
        genome_data = self.param_required("genome_data")

        # update division and url
        if "species" in genome_data:
            sd = genome_data["species"]
            # Add division, but not needed for BRC4
            if ("division" not in sd) and not brc4_mode:
                sd["division"] = division
            if "url" not in sd:
                if "production_name" in sd:
                    sd["url"] = sd["production_name"].capitalize()

        # RETROCOMPATIBILITY
        # provider name and url to assembly.provider_*: retrocompatibility
        if "provider" in genome_data:
            prov = genome_data["provider"]
            if "name" in prov and not "provider_name" in genome_data["assembly"]:
                genome_data["assembly"]["provider_name"] = prov["name"]
            if "url" in prov and not "provider_url" in genome_data["assembly"]:
                genome_data["assembly"]["provider_url"] = prov["url"]
            del genome_data["provider"]

        # BRC4 organism abbrev: move from species to BRC4 namespace
        if "species" in genome_data:
            sd = genome_data["species"]
            if "BRC4_organism_abbrev" in sd:
                if "BRC4" not in genome_data:
                    genome_data["BRC4"] = {}
                if "organism_abbrev" not in genome_data["BRC4"]:
                    genome_data["BRC4"]["organism_abbrev"] = sd["BRC4_organism_abbrev"]
                del genome_data["species"]["BRC4_organism_abbrev"]
        # END RETROCOMPATIBILITY

        # Set up display name if not set
        if "species" in genome_data:
            sd = genome_data["species"]
            if not "display_name" in sd and "scientific_name" in sd:
                genome_data["species"]["display_name"] = sd["scientific_name"]

        # flattern and dump
        flat = self.flattern(genome_data, ignore)
        flat += list(map(lambda p: (copy[p[0]], p[1]), filter(lambda x: x[0] in copy, flat)))
        flat = list(filter(lambda x: x[0] not in ignore, flat))
        if len(flat) <= 0:
            return

        # dump
        os.makedirs(wd, exist_ok=True)
        sql_file = pj(wd, "insert_meta.sql")
        with open(sql_file, "w") as sql:
            print("insert ignore into meta (species_id, meta_key, meta_value) values", file=sql)
            c = ""
            for _key, _val in flat:
                if isinstance(_val, bool):
                    _val = int(_val)
                if isinstance(_val, str):
                    _val = '"%s"' % (_val)
                print('%s (1, "%s", %s)' % (c, str(_key), str(_val)), file=sql)
                c = ","
            print(";", file=sql)
        # run insert sql
        self.run_sql_req(sql_file, log_pfx=sql_file, from_file=True)

    def flattern(self, data, ignore_lst, pfx=None):
        # vectors with values only
        if isinstance(data, list):
            return [(pfx, v) for v in data]
        elif isinstance(data, dict):
            ignore_lst = frozenset(ignore_lst)
            res = []
            for k, v in data.items():
                new_pfx = ".".join(filter(lambda p: p != None, [pfx, k]))
                res.append(self.flattern(v, ignore_lst, new_pfx))
            return sum(res, [])
        return data != None and [(pfx, data)] or []

    def run_sql_req(self, sql, log_pfx, from_file=False):
        os.makedirs(dirname(log_pfx), exist_ok=True)

        sql_option = r""" -sql '{_sql}' """.format(_sql=sql)
        if from_file:
            sql_option = r""" < '{_sql}' """.format(_sql=sql)

        cmd = r"""{_dbcmd} -url "{_srv}{_dbname}" {_sql_option} > {_out} 2> {_err}""".format(
            _dbcmd="perl %s/scripts/db_cmd.pl" % os.getenv("EHIVE_ROOT_DIR"),
            _srv=self.param("dbsrv_url"),
            _dbname=self.param("db_name"),
            _sql_option=sql_option,
            _out=log_pfx + ".stdout",
            _err=log_pfx + ".stderr",
        )
        print("running %s" % (cmd), file=sys.stderr)
        return sp.run(cmd, shell=True, check=True)

    def db_string(self):
        return "-dbhost {host_} -dbport {port_} -dbuser {user_} -dbpass {pass_} -dbname {dbname_} ".format(
            host_=self.param("dbsrv_host"),
            port_=self.param("dbsrv_port"),
            user_=self.param("dbsrv_user"),
            pass_=self.param("dbsrv_pass"),
            dbname_=self.param("db_name"),
        )

db_string()

Source code in src/python/ensembl/brc4/runnable/fill_metadata.py
135
136
137
138
139
140
141
142
def db_string(self):
    return "-dbhost {host_} -dbport {port_} -dbuser {user_} -dbpass {pass_} -dbname {dbname_} ".format(
        host_=self.param("dbsrv_host"),
        port_=self.param("dbsrv_port"),
        user_=self.param("dbsrv_user"),
        pass_=self.param("dbsrv_pass"),
        dbname_=self.param("db_name"),
    )

flattern(data, ignore_lst, pfx=None)

Source code in src/python/ensembl/brc4/runnable/fill_metadata.py
104
105
106
107
108
109
110
111
112
113
114
115
def flattern(self, data, ignore_lst, pfx=None):
    # vectors with values only
    if isinstance(data, list):
        return [(pfx, v) for v in data]
    elif isinstance(data, dict):
        ignore_lst = frozenset(ignore_lst)
        res = []
        for k, v in data.items():
            new_pfx = ".".join(filter(lambda p: p != None, [pfx, k]))
            res.append(self.flattern(v, ignore_lst, new_pfx))
        return sum(res, [])
    return data != None and [(pfx, data)] or []

param_defaults()

Source code in src/python/ensembl/brc4/runnable/fill_metadata.py
27
28
29
30
31
32
def param_defaults(self):
    return {
        "division": "",
        "ignore": [],
        "copy": {},
    }

run()

Source code in src/python/ensembl/brc4/runnable/fill_metadata.py
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
def run(self):
    # params
    wd = self.param_required("work_dir")
    division = self.param("division")
    brc4_mode = self.param("brc4_mode")
    ignore = self.param("ignore")
    copy = self.param("copy")
    genome_data = self.param_required("genome_data")

    # update division and url
    if "species" in genome_data:
        sd = genome_data["species"]
        # Add division, but not needed for BRC4
        if ("division" not in sd) and not brc4_mode:
            sd["division"] = division
        if "url" not in sd:
            if "production_name" in sd:
                sd["url"] = sd["production_name"].capitalize()

    # RETROCOMPATIBILITY
    # provider name and url to assembly.provider_*: retrocompatibility
    if "provider" in genome_data:
        prov = genome_data["provider"]
        if "name" in prov and not "provider_name" in genome_data["assembly"]:
            genome_data["assembly"]["provider_name"] = prov["name"]
        if "url" in prov and not "provider_url" in genome_data["assembly"]:
            genome_data["assembly"]["provider_url"] = prov["url"]
        del genome_data["provider"]

    # BRC4 organism abbrev: move from species to BRC4 namespace
    if "species" in genome_data:
        sd = genome_data["species"]
        if "BRC4_organism_abbrev" in sd:
            if "BRC4" not in genome_data:
                genome_data["BRC4"] = {}
            if "organism_abbrev" not in genome_data["BRC4"]:
                genome_data["BRC4"]["organism_abbrev"] = sd["BRC4_organism_abbrev"]
            del genome_data["species"]["BRC4_organism_abbrev"]
    # END RETROCOMPATIBILITY

    # Set up display name if not set
    if "species" in genome_data:
        sd = genome_data["species"]
        if not "display_name" in sd and "scientific_name" in sd:
            genome_data["species"]["display_name"] = sd["scientific_name"]

    # flattern and dump
    flat = self.flattern(genome_data, ignore)
    flat += list(map(lambda p: (copy[p[0]], p[1]), filter(lambda x: x[0] in copy, flat)))
    flat = list(filter(lambda x: x[0] not in ignore, flat))
    if len(flat) <= 0:
        return

    # dump
    os.makedirs(wd, exist_ok=True)
    sql_file = pj(wd, "insert_meta.sql")
    with open(sql_file, "w") as sql:
        print("insert ignore into meta (species_id, meta_key, meta_value) values", file=sql)
        c = ""
        for _key, _val in flat:
            if isinstance(_val, bool):
                _val = int(_val)
            if isinstance(_val, str):
                _val = '"%s"' % (_val)
            print('%s (1, "%s", %s)' % (c, str(_key), str(_val)), file=sql)
            c = ","
        print(";", file=sql)
    # run insert sql
    self.run_sql_req(sql_file, log_pfx=sql_file, from_file=True)

run_sql_req(sql, log_pfx, from_file=False)

Source code in src/python/ensembl/brc4/runnable/fill_metadata.py
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
def run_sql_req(self, sql, log_pfx, from_file=False):
    os.makedirs(dirname(log_pfx), exist_ok=True)

    sql_option = r""" -sql '{_sql}' """.format(_sql=sql)
    if from_file:
        sql_option = r""" < '{_sql}' """.format(_sql=sql)

    cmd = r"""{_dbcmd} -url "{_srv}{_dbname}" {_sql_option} > {_out} 2> {_err}""".format(
        _dbcmd="perl %s/scripts/db_cmd.pl" % os.getenv("EHIVE_ROOT_DIR"),
        _srv=self.param("dbsrv_url"),
        _dbname=self.param("db_name"),
        _sql_option=sql_option,
        _out=log_pfx + ".stdout",
        _err=log_pfx + ".stderr",
    )
    print("running %s" % (cmd), file=sys.stderr)
    return sp.run(cmd, shell=True, check=True)