Pierre Thomson wrote:
> Bowie Bailey wrote:
>>
>> Some of the medication spams are using an obnoxious html table
>> structure that makes the contents of each cell print vertically.
>>
>> For example:
>> <table>
>> <tr>
>> <td>a d g</td>
>> <td>b e h</td>
>> <td>c f i</td>
>> <td width=100%></td>
>> <\tr>
>> </table>
>>
>> This results in:
>> a b c
>> d e f
>> g h i
>>
>> Has anyone else been having this problem? Any rules to catch
>> medication names in those types of tables?
>>
>
> Here's a simple rule I wrote a couple days ago:
>
> body PT_DRUG1 /([CVAXP] ){5}/
> describe PT_DRUG1 Drug names in table of 1-letter columns
> score PT_DRUG1 3.0
>
> It works for me, no FP's yet that I am aware of. There are also
> variants for 2-letter and 3-letter bits of the same drug names.
>
If anyone can formulate a regex to catch these letters in any order, while
avoiding a repeating sequence like "A A A A A ", it would make this a safer
rule.
Pierre
