Merge pull request #241 from lcnetdev/gurmukhi

Create Gurmukhi base and nasalization hooks.

Author: scossu

doc/hooks.md

@@ -127,7 +127,9 @@ class. This may change in the future as specific needs arise.
 
 The following members of the context object are available in all the hooks:
 
-- `ctx.src`: Source text. Read only.
+- `ctx.orig`: Original source text. Read only.
+- `ctx.src`: Source text, initially equal to `ctx.orig` but it may be modified,
+  e.g. during normalization.
 - `ctx.general`: Configuration general options.
 - `ctx.langsec`: language section (S2R or R2S) of configuration.
 - `ctx.options`: language-specific options defined in configuration and set

scriptshifter/hooks/gurmukhi/__init__.py

@@ -0,0 +1,39 @@
+from logging import getLogger
+from os import path
+from re import compile
+
+from yaml import load as yload
+try:
+    from yaml import CLoader as Loader
+except ImportError:
+    from yaml import Loader
+
+
+logger = getLogger(__name__)
+
+MOD_BASEDIR = path.dirname(__file__)
+
+with open(path.join(MOD_BASEDIR, "gurmukhi_nasalization_pre.yml")) as fh:
+    nas_config = {"initial": {}, "default": {}}
+
+    for k, v in yload(fh, Loader=Loader).items():
+        if "%" in k:
+            nas_config["initial"][compile(k.replace("%", "\\b"))] = v
+        else:
+            nas_config["default"][k] = v
+
+
+def nasalize_post_config(ctx):
+    """
+    Preprocess Roman input to get a uniform nasalization.
+
+    The result is not a correct romanization: it is an intermediate stage
+    to be passed to the Gurmukhi table.
+    """
+    for k, v in nas_config["initial"].items():
+        ctx.src = k.sub(v, ctx.src)
+    for k, v in nas_config["default"].items():
+        ctx.src = ctx.src.replace(k, v)
+
+    if ctx.orig != ctx.src:
+        logger.debug(f"Corrected nasalization: {ctx.orig} -> {ctx.src}")