Skip to content

Latest commit

 

History

History
141 lines (98 loc) · 6.94 KB

README.md

File metadata and controls

141 lines (98 loc) · 6.94 KB
Raw

LizMap import module

Presentation

LizMap module to enable the import of data from CSV files into configured PostgreSQL layers.

The map editor can activate the import tool for a layer inside a LizMap project.

When the user clicks on the configured layers in the Layers panel, a new interface will be displayed in the Information panel displayed at the right.

sub-dock interface

The user can then upload a CSV file and either:

  • validate the uploaded CSV data against a set of rules defined by the editor
  • import the data into the target layer if the data is correct.

Principle

The map editor will configure:

  • for which layer(s) the authenticated user can import data. The target layer PostgreSQL schema and table must be specified.
  • which group(s) of users can access the tool in the map interface.
  • which fields can be imported in the target table
  • if the CSV contains geometry data: then two fields longitude and latitude must be present in the CSV file. (At present, this tool only allows the import of geometries for point data)
  • the set of rules that define a valid record. Each rule will be tested for all the CSV features.
  • The name of the mandatory unique ID field to allow identifying each record: the tool will check that this field is present and that its values are all unique.

If an error is detected (wrong CSV format, missing fields, missing unique id field, etc.) the tool will cancel the import and display an error message.

If some features do not respect the rules, a table will show the list of rules with the problematic identifiers.

sub-dock interface

If some records contained in the CSV files are duplicates of some target layers records (in the PostgreSQL table), the whole import will be canceled. A list of the duplicated records identifiers will be displayed.

sub-dock interface

After a successful import, the target layer will be refreshed in the map.

The imported data will have a new import_metadata JSON field containing the metadata of the import: the user login, the import date and time.

{
    "import_login": "admin",
    "import_temp_table": "temp_1674060987_target",
    "import_time": "2023-01-18T16:56:27"
}

Installation

See Installation instrutions

Configuration

The configuration of the target layers and of the rules are stored inside a PostgreSQL database (the one containing the target layers tables).

The import module is based on :

  • Two tables containing the list of target layers and the list of rules
  • PostgreSQL Functions used by the module to:
    • import the data into temporary tables
    • check that the defined rules are respected by the data
    • verify if there are duplicated data between the source CSV data and the target table
    • proceed and insert the data into the target table.
  • A schema lizmap_import_module containing these tables and functions

Tables

There are 2 tables used to store the module configuration:

  • lizmap_module_import.import_csv_destination_tables: the list of target layers
  • lizmap_module_import.import_csv_field rules: the list of rules for each target layer

Destination tables

The table lizmap_module_import.import_csv_destination_tables contains the list of target layers. You need to define:

  • a PostgreSQL table schema & name which define the layer data source,
  • a LizMap repository key and project key which define on which LizMap map to activate the import tool,
  • the target fields expected to be in the CSV, written as a PostgreSQL array of text text[],
  • the geometry source (at present, only lonlat is possible but wkt will be supported in future versions.)

Example content:

id table_schema table_name lizmap_repository lizmap_project target_fields geometry_source unique_id_field
1 demo trees tests import {height,genus,leaf_type} lonlat id_csv

Field rules

The table lizmap_module_import.import_csv_field_rules contains the list of the rules.

These rules are made to control that the data to import from the CSV file respect some pre-defined conditions.

You can have more than one line per target layer. The target layer is defined by its target_table_schema and target_table_name.

The criteria_type column must only accept values among:

  • not_null: the rule defines that the field must not be empty (or NULL)
  • format: the rule checks if the given value for the field is in the desired format (integer, real, text, uuid, boolean, date, time, timestamp)
  • valid: the rule refers to a specific condition.

The code column is made to store a simple name for your rule. The label must be used to describe the rule (the label will be displayed in the report)

The SQL condition must be entered in the condition column. For the format criteria type you can use the function lizmap_import_module.import_csv_is_given_type(text, text) as shown in the examples below.

Example content:

id target_table_schema target_table_name criteria_type code label description condition join_table
1 demo trees not_null genus_not_null The field genus cannot be empty genus IS NOT NULL
2 demo trees not_null leaf_type_not_null The field leaf_type cannot be empty leaf_type IS NOT NULL
3 demo trees format height_format The field height must be a real number lizmap_import_module.import_csv_is_given_type(height, 'integer')
4 demo trees valid height_valid The height value must be between 1.0 and 30.0 height BETWEEN 1.0 AND 30.0
5 demo trees valid genus_valid The genus must be Platanus or Cupressus "genus IN ('Cupressus', 'Platanus')"