Main trapnakron constructor, covering all options.

The trapnakron constructor sets, either by default or optionally, various transformations to enhance queries to the resulting derivator.

Default Behavior

Property values are returned as alternatives/hybridized compositions of Ekavian Cyrillic, Ekavian Latin, Ijekavian Cyrillic, and Ijekavian Latin forms, as applicable. Any of these forms can be excluded from derivation by setting its env* parameter to None. env* parameters can also be used to change the priority environment from which the particular form is derived.

Derivation and property key separator in compound keys is the ASCII hyphen (-).

Derivation keys are derived from syntagmas by applying the identify() function. In derivations where this will result in strange keys, additional keys should be defined through hidden syntagmas. Property keys are transliterated into stripped-ASCII.

Conflict resolution for derivation keys is not strict (see derivator constructor).

Optional behavior

Instead of plain text, properties may be reported with some markup. The markup type is given by markup parameter, and can be one of "plain", "xml", "docbook4". The tagmap parameter contains mapping of derivation keys to tags which should wrap properties of these derivations.

Derivation keys can have several suffixes which effect how the properties are reported:

• Presence of the suffix given by ptsuff parameter signals that properties should be forced to plain text, if another markup is globally in effect.
• Parameter ltsuff states the suffix which produces lighter version of the markup, where applicable (e.g. people names in Docbook).
• When fetching a property within a sentence (with keys given e.g. as XML entities), sentence construction may require that the resolved value is of certain gender and number; parameter gnsuff can be used to provide a tuple of 4 suffixes for gender in singular and 4 suffixes for gender in plural, such that the property will resolve only if the value of gender and number matches the gender and number suffix.
• Parameters stsuff and adsuff provide suffixes through which systematic transcription and alternative derivations are requested. They are actually tuples, where the first element is the key suffix, and the second element the suffix to primary environment which produces the systematic/alternative environment. adsuff can also be a tuple of tuples, if several alternative derivations should be reachable.
• In case the entry is a person's name with tagged first and last name, parameter nmsuff can provide a tuple of 2 suffixes by which only the first or last name are requested, respectively.

Ordinary hyphens may be converted into non-breaking hyphens by setting the nobrhyp parameter to True. Non-breaking hyphens are added heuristically, see the to_nobr_hyphens() hook. Useful e.g. to avoid wrapping on hyphen-separated case endings.

A property key normally cannot be empty, but npkeyto parameter can be used to automatically substitute another property key when empty property key is seen in request for properties. In the simpler version, value of npkeyto is just a string of the key to substitute for empty. In the more complex version, the value is a tuple containing the key to substitute and the list of two or more supplemental property keys: empty key is replaced only if all supplemental property values exist and are equal (see e.g. trapnakron_plain for usage of this).

Some property values may have been manually decorated with disambiguation markers (¤), to differentiate them from property values of another derivation which would otherwise appear equal under a certain normalization. By default such markers are removed, but instead they can be substituted with a string given by disamb parameter.

Some derivations are defined only for purposes of obtaining their properties in scripted translations at runtime. They are by default not included, but can be by setting the runtime parameter to True.

Parameters:

• envec (string or None) - primary environment for Ekavian Cyrillic derivation
• envel (string or None) - primary environment for Ekavian Latin derivation
• envic (string or None) - primary environment for Ijekavian Cyrillic derivation
• envil (string or None) - primary environment for Ijekavian Latin derivation
• markup (string) - target markup
• tagmap (dict string -> string) - tags to assign to properties by derivation keys
• ptsuff (string) - derivation key suffix to report plain text properties
• ltsuff (string) - derivation key suffix to report properties in lighter markup
• gnsuff ([(string, string)*]) - suffixes by gender and number, to have no resolution if gender or number do not match
• stsuff ((string, string)) - derivation key and environment name suffixes to report systematic transcriptions
• adsuff ((string, string) or ((string, string)*)) - derivation key and environment name suffixes to report alternative derivations
• nmsuff ((string, string)) - suffixes for fetching only first or last name of a person
• npkeyto (string or (string, [string*])) - property key to substitute for empty key, when given
• nobrhyp (bool) - whether to convert some ordinary into non-breaking hyphens
• disamb (string) - string to replace each disambiguation marker with
• runtime (bool) - whether to include runtime-only derivations

Returns: Synder

trapnakron derivator

Get root directory to trapnakron derivation files.

Returns: string

root directory path

Constructs trapnakron suitable for application to plain text.

Calls trapnakron with the following setup:

• Markup is plain text (plain).
• Suffixes: _rm ("rod muski") for resolving the property value only if it is of masculine gender, _rz for feminine, _rs for neuter; _s for systematic transcription, _a, _a2 and _a3 for other alternatives; _im and _pr for person's last and first name.
• Non-breaking hyphens are heuristically replacing ordinary hyphens.
• Empty property key is converted into am (accusative masculine descriptive adjective), providing that it is equal to gm (genitive masculine descriptive adjective); i.e. if the descriptive adjective is invariable.

Constructs trapnakron suitable for application to UI texts.

Like trapnakron_plain, except that disambiguation markers are not removed but substituted with an invisible character, and runtime-only derivations are included too.

Retaining disambiguation markers is useful when a normalized form (typically nominative) is used at runtime as key to fetch other properties of the derivation, and the normalization is such that it would fold two different derivations to same keys if the originating forms were left undecorated.

Constructs trapnakron suitable for application to Docbook 4 texts.

Calls trapnakron with the following setup:

• Markup is Docbook 4 (docbook4).
• Suffixes: _ot ("obican tekst") for plain-text properties, _lv ("laksa varijanta") for lighter variant of the markup. Lighter markup currently applies to: people names (no outer <personname>, e.g. when it should be elideded due to particular text segmentation on Docbook->PO extraction). Also the suffixes as for trapnakron_plain.
• Non-breaking hyphens and empty property keys are treated like in trapnakron_plain.

Normalize internal property keys in trapnakron.

Parameters:

• pkey (string or (string*) or [string*]) - property key or keys to normalize

Returns: as input

normalized keys

Normalize text into runtime key for translation scripting.

Parameters:

• text (string) - text to normalize into runtime key

Returns: string

runtime key