prep - Prepare Collations for Stemma

Prep is a command-line utility written in C that preprocesses textual collation data for use with the stemma program. It processes a collation file containing manuscript variant information and generates three output files: a matrix of taxa and variants (.tx), stratigraphical constraints (.no, and a listing of each variant (.vr). These outputs are tailored for stemmatic analysis, enabling the stemma program to infer relationships between manuscripts.

Overview

In textual criticism, a collation compares different manuscripts or versions of a text to identify variations. Prep takes such collation data, optionally filters it based on user-defined criteria (e.g., fragmentary witnesses, corrected manuscripts, or chronological constraints), and transforms it into structured formats suitable for phylogenetic or stemmatic analysis. Usage

bash
prep collation_file [taxa...]

• collation_file: The input file containing collation data (typically with no extension). ** taxa...: An optional space-separated list of specific manuscript names (taxa) or macros defined in the file to include in the output. If omitted, all witnesses are processed.

Example

Preprocess a collation file named mss:

bash
./prep mss

Preprocess only specific manuscripts:

bash
./prep mss MS1 MS2 MS3

Input Format

The collation file uses a specific syntax with tokens to define manuscripts, variants, and metadata. Key tokens include:

• *: Defines the semicolon terminated list of manuscripts (witnesses), e.g., * MS1 MS2 MS3 ;.
• /: Switches between parallel texts, e.g., /a for parallel 'a'. Useful for Synoptic gospels.
• =: Defines or modifies macros (for sets of manuscripts), e.g., = $m MS1 MS2 ; sets macro $m.
• %: Manages lacunae (gaps), e.g., %- MS1 ; marks MS1 as lacunose starting from that point in the text.
• @: Marks a verse or section, e.g., @ John 1:1.
• [: Begins a list of readings (variants), e.g., [ lemma | reading1 | reading2 ].
• <: Begins a list of witnesses for readings, e.g., < 00 MS1 MS2 | 11 MS3 > (NB: the number of digits in the first item must equal the number of readings in the previous list of readings).
• ^: Specifies a chronological file, e.g., ^ chron.txt.
• ~: Defines aliases, e.g., ~ MS1 GA123 PrintName (useful if your collation with one notation and your chron information in another).
• ": Encloses double-quotation terminated comments, e.g., " This is a comment ".
• -: Removes collated manuscripts from the output, e.g., - MS1 ;.
• +: Skips tokens until the semicolon terminator, e.g., + junk ; (useful for commenting out macro definitions, etc.). For detailed syntax, refer to the comments in prep.c under the Tokens: section.

Sample Input

* MS1 MS2 MS3 ;
@ John 1:1
[ In principio | erat | fuit ]
< 0 MS1 MS2 | 1 MS3 >

Output Files

Prep generates three files from the input collation:

1. mss.tx:

• A matrix where rows represent taxa (manuscripts) and columns represent weighted variants.
• Each cell indicates the variant state (e.g., '0', '1', or '?') for a manuscript.
• Format: <number of witnesses> <number of weighted variants> followed by rows of witness names and variant states.

1. mss.no:

• Stratigraphical constraints based on chronological data (if provided via a ^ token).
• Lists each witness with its stratum (time layer) and manuscripts it postdates.
• Format: <witness> <stratum> < <preceding witnesses> >.

1. mss.vr: A listing of each variant unit, including the lemma and possible readings. Format: Verse markers (@) followed by lemma and numbered readings (e.g., > In principio\n1=erat 2=fuit). These files are designed as inputs for the Stemma program for further analysis.

Environmental Variables

Customize Prep's behavior with environment variables (set via export in bash or prefixed on the command line calling prep):

• YEARGRAN: Controls year granularity for stratigraphy.
  • Default: 0 (no granularity).
  • -1: Tuned for New Testament manuscripts (uses a predefined table).
  • Positive integer: Groups years into specified intervals.
• FRAG:
  • Threshold for including fragmentary witnesses (as a percentage or number of non-constant variants).
  • Default: 50%.
  • Example: export FRAG=75% or export FRAG=10.
• CORR:
  • Threshold for including corrected witnesses (as a percentage or number of new variants).
  • Default: 10%.
  • Example: export CORR=20%.
• YEAR:
  • Cutoff year; excludes witnesses dated after this year.
  • Default: No cutoff.
  • Example: export YEAR=1200.
• NOSING:
  • If set (e.g., export NOSING=1), excludes singular readings (variants unique to one witness) from the matrix.
• ROOT:
  • Defines an explicit root or ancestor (e.g., the archetype or a standard text).
  • Default: No root.
  • Example: export ROOT=UBS.
• KEEPSAME:
  • Turns off the identical witness check.
  • Default: suppress identical witnessesd
  • Example: export KEEPSAME

Special Macros

• $*: Represents all active witnesses.
• $?: Represents witnesses with unknown readings (e.g., lacunae or missing data).

License

This project is licensed under the MIT License. See the LICENSE file for details.

Acknowledgements

• The prep program is designed to work with the stemma program for stemmatic analysis.