# Basic instructions for running the LdmlConverter via Maven > Note: While this document provides useful background information about the `LdmlConverter`, the actual complete process for integrating CLDR data to ICU is described in the document `../../../docs/processes/cldr-icu.md` which is best viewed as [CLDR-ICU integration](https://unicode-org.github.io/icu/processes/cldr-icu.html) ## TLDR * Define the `ICU_DIR`, `CLDR_DIR`, and `CLDR_DATA_DIR` environment variables, or (see below) * Check / update versions * Build ICU4J: ```sh cd "$ICU_DIR" mvn clean install -f icu4j -DskipTests -DskipITs ``` * Build the `cldr-code` library from the `cldr` repo: ```sh cd "$CLDR_DIR" mvn clean install -pl :cldr-all,:cldr-code -DskipTests -DskipITs ``` * Build the conversion tool: ```sh cd "$ICU_DIR/tools/cldr/cldr-to-icu/" mvn clean package -DskipTests -DskipITs ``` * Run the conversion tool: ```sh java -jar target/cldr-to-icu-1.0-SNAPSHOT-jar-with-dependencies.jar ``` ## Requirements * A CLDR release for supplying CLDR data and the CLDR API. * JDK 11+ * The Maven build tool ## Important directories | Directory | Description | |-----------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | `ICU_DIR` | Path to root of ICU directory, below which are (e.g.) the `icu4c/`, `icu4j/` and `tools/` directories. | | `CLDR_DIR` | This is the path to the to root of standard CLDR sources, below which are the `common/` and `tools/` directories. | | `CLDR_DATA_DIR` | The top-level directory for the CLDR production data (typically the "production" directory in the staging repository). Usually generated locally or obtained from: https://github.com/unicode-org/cldr-staging/tree/main/production | In Posix systems, it's best to set these as exported shell variables, and any following instructions assume they have been set accordingly: ```sh export TOOLS_ROOT=/path/to/icu/tools export CLDR_DIR=/path/to/cldr export CLDR_DATA_DIR=/path/to/cldr-staging/production ``` Note that you should not attempt to use data from the CLDR project directory (where the CLDR API code exists) for conversion into ICU data. The process now relies on a pre-processing step, and the CLDR data must come from the separate "staging" repository (i.e. https://github.com/unicode-org/cldr-staging) or be pre-processed locally into a different directory. :point_right: **Note**: the 3 folders can also be overridden: * with Java properties (e.g. `-DCLDR_DIR=/foo/bar`) * from the command line when invoking the tool (the `icuDir`, `cldrDir`, and `cldrDataDir` options) ## Initial Setup This project relies on the Maven build tool for managing dependencies, so it will need to be installed. On a Debian based system, this should be as simple as: ```sh sudo apt-get install maven ``` ## Check / update versions ### Real versions **ICU version (`real_icu_ver`):** ```sh mvn help:evaluate -Dexpression=project.version -q -DforceStdout -f $ICU_DIR/icu4j ``` **CLDR Library version (`real_cldr_ver`):** ```sh mvn help:evaluate -Dexpression=project.version -q -DforceStdout -f $CLDR_DIR/tools ``` ### Dependency versions **ICU version used by the cldr conversion tool:** \ ⚠️ **Warning:** Must be the same as `real_icu_ver` ```sh mvn help:evaluate -Dexpression=icu4j.version -q -DforceStdout -f $ICU_DIR/tools/cldr/cldr-to-icu ``` **CLDR library version used by the cldr conversion tool:** \ ⚠️ **Warning:** Must be the same as `real_cldr_ver` ```sh mvn help:evaluate -Dexpression=cldr-code.version -q -DforceStdout -f $ICU_DIR/tools/cldr/cldr-to-icu ``` **ICU version used by the cldr library:** \ ⚠️ **Warning:** Must be the same as `real_icu_ver` ```sh mvn help:evaluate -Dexpression=icu4j.version -q -DforceStdout -f $CLDR_DIR/tools ``` ### TLDR (Quick update versions without checking) ```sh # Get real versions real_icu_ver=`mvn help:evaluate -Dexpression=project.version -q -DforceStdout -f $ICU_DIR/icu4j` echo $real_icu_ver real_cldr_ver=`mvn help:evaluate -Dexpression=project.version -q -DforceStdout -f $CLDR_DIR/tools` echo $real_cldr_ver # Set dependency versions mvn versions:set-property -DgenerateBackupPoms=false -Dproperty=icu4j.version -DnewVersion=$real_icu_ver -f $ICU_DIR/tools/cldr/cldr-to-icu mvn versions:set-property -DgenerateBackupPoms=false -Dproperty=cldr-code.version -DnewVersion=$real_cldr_ver -f $ICU_DIR/tools/cldr/cldr-to-icu mvn versions:set-property -DgenerateBackupPoms=false -Dproperty=icu4j.version -DnewVersion=$real_icu_ver -f $CLDR_DIR/tools ``` ## Build everything You must also build and install an additional CLDR library in the the local Maven repository. Since that depends on ICU4J, you need to build and install that first. Lastly, build the conversion tool ```sh # Build ICU4J cd "$ICU_DIR" mvn clean install -f icu4j -DskipTests -DskipITs # Build the CLDR library cd "$CLDR_DIR" mvn clean install -pl :cldr-all,:cldr-code -DskipTests -DskipITs # Build the conversion tool cd "$ICU_DIR/tools/cldr/cldr-to-icu/" mvn clean package -DskipTests -DskipITs ``` ## Generating all ICU data and source code Run the conversion tool: ```sh cd "$ICU_DIR/tools/cldr/cldr-to-icu/" java -jar target/cldr-to-icu-1.0-SNAPSHOT-jar-with-dependencies.jar ``` You can run it with `--help` for all the options supported. ## Other Examples * Outputting a subset of the supplemental data into a specified directory: ```sh java -jar target/cldr-to-icu-1.0-SNAPSHOT-jar-with-dependencies.jar --outDir=/tmp/cldr --outputTypes=plurals,dayPeriods --dontGenCode ``` Note: Output types can be listed with mixedCase, lower_underscore or UPPER_UNDERSCORE. Pass `-DoutputTypes=help` to see the full list. * Outputting only a subset of locale IDs (and all the supplemental data): ```sh java -jar target/cldr-to-icu-1.0-SNAPSHOT-jar-with-dependencies.jar --outDir=/tmp/cldr --outputTypes=plurals,dayPeriods --dontGenCode java -jar target/cldr-to-icu-1.0-SNAPSHOT-jar-with-dependencies.jar --outDir=/tmp/cldr --localeIdFilter='(zh|yue).*' --dontGenCode ``` * Overriding the default CLDR version string (which normally matches the CLDR library code): ```sh java -jar target/cldr-to-icu-1.0-SNAPSHOT-jar-with-dependencies.jar --cldrVersion="36.1" ``` ### Using `alt="ascii"` CLDR alternate values from the CLDR XML CLDR provides alternate values in addition to the default values for locale data. For example, some locales have time formats using U+202F NARROW NO-BREAK SPACE (`NNBSP`) between the hours/minutes/seconds and the day periods. In order to provide the equivalent time formats that use the ASCII space U+0020 SPACE, the alternate values have the extra attribute `alt="ascii"`. Follw these steps to generate ICU data using the ASCII versions of locale data: 1. First, edit the `config.xml` file where it mentions `ALTERNATE VALUES` with the correctly annotated source path, target path, and locales list as follows: ```diff @@ -384,6 +399,20 @@ + + + + + + + + + + + + + + + + ``` 1. Then run the generator: ```sh java -jar target/cldr-to-icu-1.0-SNAPSHOT-jar-with-dependencies.jar ``` ## Config syntax details Note: some elements have an implicit default attributes associated with them, according to [`ldml.dtd`](https://github.com/unicode-org/icu/blob/main/icu4c/source/data/dtd/cldr/common/dtd/ldml.dtd). For example, for the `timeFormat` element, the following excerpt of the DTD schema indicates that there is a default value `"standard"` for the `type` attribute: ``` ``` See `config.xml` for documentation of all options and additional customization. ## Running unit tests (CURRENTLY FAILING) ```sh mvn test -DCLDR_DIR="$CLDR_DATA_DIR" ``` ## Importing and running from an IDE This project should be easy to import into an IDE which supports Maven development, such as IntelliJ or Eclipse. It uses a local Maven repository directory for the unpublished CLDR libraries (which are included in the project), but otherwise gets all dependencies via Maven's public repositories. But before importing and running it you still need to build the ICU4J and the CLDR library (see above).