Customizing the fexcepts.txt table

The 'fexcept.txt' table is a small file distributed with MARC Report that can be customized to 'fine-tune' warning messages for initial articles.

fexcept.txt is a plain-text file consisting of four-coulmn rows in tab-delimited format¹⁾. Here are the first three rows of this table:

Y	A	fre	Y
Y	A	spa	Y
Y	A	ita	Y

The definitions of each column are as follows:

• Col.1: Enable/Disable the exception: 'Y' to enable, 'N' to disable
• Col.2: Initial article: must be in uppercase
• Col.3: Language code: must be in lowercase
• Col.4: Also apply to titles in subfield $t: 'Y' to enable, 'N' to disable

How it works

In a MARC record, if a normalized article in a title field matches an article in column 2, and the lang code from the record's 008 matches the corresponding code in column 3, and the exception is enabled (Column 1 = 'Y'), then the MARC Report warning message is suppressed.

In addition to this, there is one special option that can be applied: To suppress all warnings for a particular article, set the lang code to 'any', and delete all other entries in the table (if applicable) for that article.

Therefore, the default MARC Report behavior for initial articles is to warn the user whenever:

1. the first word of a title is an initial article in any language, and
2. the non-filing indicator does not match the number of characters in the initial article, and
3. there is no exception for that initial article in the fexcepts.txt table.

An example case

Here is an example which should illustrate why this table is needed and how it can be customized. Consider a title like:

245 10 $aI am a Jedi apprentice

'I' is an initial article only in Italian and Welsh. MARC Report knows about titles, and it knows about initial articles–what it does not know about is what language any particular title is rendererd in. So, when the program sees the 'I ' at the beginning of a title, the default behavior is to show the message re: '245: Check language code'²⁾.

But, as some initial articles are common beginning words in English titles (like 'I', 'DIE', 'TEN', etc.), these messages may become a nuisance and be cancelled or turned off. This might be likely when catalogers are trying to quickly validate records and do not want to be concerned with false alerts.

However, if disabled, these messages will not show up in those cases where they might be useful.

It is for this scenario that the fexcepts.txt table was designed. It was not designed to be a finished product, but more a starting point for the user. To create it, we looked at the LC database and tried to identify the most common false alerts for initial articles, and then compiled the results into this table.

Thus, the row:

Y     I     eng   Y

cancels any of the above warning messages about 'I' being an initial article whenever the language code is English. If you are also cataloging materials in Welsh, then you may also wish to addd the row:

Y     I     wel   Y

to the table.

Thus, instead of simply cancelling this message–forever–by customizing a Cataloging Check set³⁾, we can cancel it selectively by adding rows to the fexcepts.txt file.