view load_db.py @ 0:3708ff0198b7 draft

planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/query_tabular commit 74915fc9cee746bbce1c4b507e13231259de177d
author iuc
date Tue, 18 Jul 2017 09:07:07 -0400
parents
children 8a33b442ecd9
line wrap: on
line source

#!/usr/bin/env python

from __future__ import print_function

import sys

from filters import TabularReader


def getValueType(val):
    if val or 0. == val:
        try:
            int(val)
            return 'INTEGER'
        except:
            try:
                float(val)
                return 'REAL'
            except:
                return 'TEXT'
    return None


def get_column_def(file_path, table_name, skip=0, comment_char='#',
                   column_names=None, max_lines=100, load_named_columns=False,
                   filters=None):
    col_pref = ['TEXT', 'REAL', 'INTEGER', None]
    col_types = []
    col_idx = None
    try:
        tr = TabularReader(file_path, skip=skip, comment_char=comment_char,
                           col_idx=None, filters=filters)
        for linenum, fields in enumerate(tr):
            if linenum > max_lines:
                break
            try:
                while len(col_types) < len(fields):
                    col_types.append(None)
                for i, val in enumerate(fields):
                    colType = getValueType(val)
                    if col_pref.index(colType) < col_pref.index(col_types[i]):
                        col_types[i] = colType
            except Exception as e:
                print('Failed at line: %d err: %s' % (linenum, e),
                      file=sys.stderr)
    except Exception as e:
        print('Failed: %s' % (e), file=sys.stderr)
    for i, col_type in enumerate(col_types):
        if not col_type:
            col_types[i] = 'TEXT'
    if column_names:
        col_names = []
        if load_named_columns:
            col_idx = []
            for i, cname in enumerate(
                    [cn.strip() for cn in column_names.split(',')]):
                if cname != '':
                    col_idx.append(i)
                    col_names.append(cname)
            col_types = [col_types[i] for i in col_idx]
        else:
            col_names = ['c%d' % i for i in range(1, len(col_types) + 1)]
            for i, cname in enumerate(
                    [cn.strip() for cn in column_names.split(',')]):
                if cname and i < len(col_names):
                    col_names[i] = cname
    else:
        col_names = ['c%d' % i for i in range(1, len(col_types) + 1)]
    col_def = []
    for i, col_name in enumerate(col_names):
        col_def.append('%s %s' % (col_names[i], col_types[i]))
    return col_names, col_types, col_def, col_idx


def create_table(conn, file_path, table_name, skip=0, comment_char='#',
                 pkey_autoincr=None, column_names=None,
                 load_named_columns=False, filters=None,
                 unique_indexes=[], indexes=[]):
    col_names, col_types, col_def, col_idx = \
        get_column_def(file_path, table_name, skip=skip,
                       comment_char=comment_char, column_names=column_names,
                       load_named_columns=load_named_columns, filters=filters)
    col_func = [float if t == 'REAL' else int
                if t == 'INTEGER' else str for t in col_types]
    table_def = 'CREATE TABLE %s (\n    %s%s\n);' % (
                table_name,
                '%s INTEGER PRIMARY KEY AUTOINCREMENT,' %
                pkey_autoincr if pkey_autoincr else '',
                ', \n    '.join(col_def))
    # print >> sys.stdout, table_def
    insert_stmt = 'INSERT INTO %s(%s) VALUES(%s)' % (
                  table_name, ','.join(col_names),
                  ','.join(["?" for x in col_names]))
    # print >> sys.stdout, insert_stmt
    data_lines = 0
    try:
        c = conn.cursor()
        c.execute(table_def)
        conn.commit()
        c.close()
        for i, index in enumerate(unique_indexes):
            index_name = 'idx_uniq_%s_%d' % (table_name, i)
            index_columns = index.split(',')
            create_index(conn, table_name, index_name, index_columns,
                         unique=True)
        for i, index in enumerate(indexes):
            index_name = 'idx_%s_%d' % (table_name, i)
            index_columns = index.split(',')
            create_index(conn, table_name, index_name, index_columns)
        c = conn.cursor()
        tr = TabularReader(file_path, skip=skip, comment_char=comment_char,
                           col_idx=col_idx, filters=filters)
        for linenum, fields in enumerate(tr):
            data_lines += 1
            try:
                vals = [col_func[i](x)
                        if x else None for i, x in enumerate(fields)]
                c.execute(insert_stmt, vals)
            except Exception as e:
                print('Failed at line: %d err: %s' % (linenum, e),
                      file=sys.stderr)
        conn.commit()
        c.close()
    except Exception as e:
        exit('Error: %s' % (e))


def create_index(conn, table_name, index_name, index_columns, unique=False):
    index_def = "CREATE %s INDEX %s on %s(%s)" % (
                'UNIQUE' if unique else '', index_name,
                table_name, ','.join(index_columns))
    c = conn.cursor()
    c.execute(index_def)
    conn.commit()
    c.close()