Awesome, thank you very much. Especially a big +1 for the 'UNION ALL' 
directive! :)


On Wednesday, 4 January 2017 16:54:47 UTC-5, Adam Cox wrote:
>
> This is what Lucy was so graciously referring to: 
> https://arches-hip.readthedocs.io/en/latest/extra/#from-ms-excel-workbook
>  :)
>
> You are most of the way there, but pay close attention to the GROUPID. 
> Technically what you have written will work fine, but if you try to add two 
> attributes to the same resource that should be grouped (like name and name 
> type, as is illustrated in the documentation above) you will need to set 
> the groupids for those resources to match.
>
> As for relationships, that will be tricky. In the past, I've written small 
> python scripts to read from tables and create the .relations file. 
> Generally these scripts are very specific though, so not really applicable 
> outside of their initial use, and I haven't used SQL for that at all. Note 
> that you will need to use the RESOURCEIDs from your .arches file in the 
> .relations file.
>
> Perhaps others have some good SQL samples for achieving this task.
>
> Good luck! Migrating data is often a very complex task, so I'm interested 
> to hear how it goes for you.
>
> On Wed, Jan 4, 2017 at 1:46 PM, Vincent Meijer <[email protected] 
> <javascript:>> wrote:
>
>> Where can I find Adam Cox's excellent guide to creating a .Arches file? :)
>>
>> I am importing an old database that is a huge mess, and it seems I will 
>> need to prepare a .arches file by querying every column of every table into 
>> a .arches format. 
>> E.g. this query:
>>
>> select CONCAT('HERITAGE_RESOURCE_', SpecimenID) as "RESOURCEID", 
>> 'HERITAGE_RESOURCE.E18' as "RESOURSETYPE", 
>> 'NAME.E41' as "ATTRIBUTENAME", 
>> CommonName as "ATTRIBUTEVALUE", 
>> 'NAME.E41' as "GROUPID"
>> from dbo.Specimens;
>>
>> will result in: 
>>
>> RESOURCEID   RESOURCETYPE ATTRIBUTENAME    ATTRIBUTEVALUE  GROUPID
>> HERITAGE_RESOURCE_4  HERITAGE_RESOURCE.E18    NAME.E41    Hot Springs 
>> NAME.E41
>> HERITAGE_RESOURCE_5   HERITAGE_RESOURCE.E18 NAME.E41     Sapsuk NAME.E41
>> HERITAGE_RESOURCE_6   HERITAGE_RESOURCE.E18 NAME.E41     Sapsuk NAME.E41
>> HERITAGE_RESOURCE_7   HERITAGE_RESOURCE.E18 NAME.E41    Adamagan NAME.E41
>>
>>
>>
>> And then of course there are the relations...
>>
>> Any advise/tactics/example sql queries are very welcome! :D
>>
>>
>>
>> On Wednesday, 17 February 2016 12:46:48 UTC-5, Adam Cox wrote:
>>>
>>> Hi Lucy, I have come to a similar conclusion in the past.  It's been 
>>> beneficial to batch import a single .arches file with a good deal of 
>>> information in it, but with no attempt to get absolutely everything.  
>>> Programming a way to convert some of the extra random pieces to .arches 
>>> seemed too costly (time-costly) compared to the relative ease of entering 
>>> that information through the arches interface.  However, populating the 
>>> entire database with a core of base data that could be added to later is 
>>> extremely valuable.
>>>
>>> Also, heads-up!  In v4 it looks like 
>>> <https://github.com/archesproject/arches/wiki/Import-Export> the 
>>> .arches format will give way to a more straightforward import approach.  
>>> That won't be relevant to you for a while, but just something to keep in 
>>> mind before you spend extra time building processes for .arches that you 
>>> expect to use for years...
>>>
>>> Adam
>>>
>>> On Wed, Feb 17, 2016 at 4:07 AM, Lucinda Fletcher-Jones <
>>> [email protected]> wrote:
>>>
>>>> Hi Adam,
>>>>
>>>> Your help is always appreciated and I really like the MS Excel query 
>>>> method of converting the excel file into an .arches file but that 
>>>> perennial 
>>>> problem, diacritics, is rearing its ugly head again. The MS query cannot 
>>>> import the diacritics from the Excel file (I have tried many times using 2 
>>>> different versions of Excel) and so adding those extra rows as you suggest 
>>>> below would be necessary for every resource as we have the Arabic name and 
>>>> an English name with diacritics for each. It probably wouldn't take that 
>>>> long but allows for human error to creep in...
>>>>
>>>> I also looked into using Open Office. By googling, I found a method by 
>>>> which I could use OO Base, import the Excel file and create a query from 
>>>> that. But, the modified query is not working at present. I have some 
>>>> syntax 
>>>> problems but I also believe that this method can only handle very simple 
>>>> queries.
>>>> As for the macros, I don't know python and my Javascript is rusty, so I 
>>>> might leave that to a new programmer coming on board soon. 
>>>> I am rapidly coming to the conclusion that a combination of batch input 
>>>> through .arches files and user input might be best for our project!
>>>>
>>>>
>>>>
>>>>
>>>> On Monday, February 15, 2016 at 7:42:42 PM UTC+2, Adam Cox wrote:
>>>>>
>>>>> Hi Lucy, personally I'd recommend modifying (a copy of!) the excel 
>>>>> file, because then you can continually recreate the .arches file from 
>>>>> it--I've always found it beneficial to spend the most time on replicable 
>>>>> steps rather than one time (copy/paste) operations.
>>>>>
>>>>> Here's something that could help the process, and I'll use the Name, 
>>>>> Name Type as an example:
>>>>>
>>>>> If you have a column of names, and you know that they are all primary 
>>>>> names, you can avoid creating a new column where every value is 
>>>>> "NAME_TYPE:1" by just hard-coding that term into a statement in the 
>>>>> query.  
>>>>> So, your new statement would not reference a column name, but would just 
>>>>> have "NAME_TYPE:1" written into it.  In the example I made, the geometry 
>>>>> column shows an example of combining strings, 'POINT (', with column 
>>>>> names, lat.  So that may be helpful for reference.
>>>>>
>>>>> A couple of other points related to this:
>>>>>
>>>>> If you have begun to use open office, it would be really great to see 
>>>>> if this same SQL stuff can be used there.  Just a word of warning, even 
>>>>> though it's all SQL, the syntax (double vs. single quotes, the use of AS) 
>>>>> may differ between Excel and Open Office.
>>>>>
>>>>> Also, it occurs to me that once you have created a spreadsheet that is 
>>>>> formatted like a .arches file, you could probably do some pretty nimble 
>>>>> things with it by using the sorting and filtering capabilities that Excel 
>>>>> has and I'm sure open office has.  A .arches file is just a bunch of rows 
>>>>> with two levels of sorting: first by RESOURCEID and second by GROUPID.  
>>>>> To 
>>>>> add a line to a specific resource, you could just filter the rows based 
>>>>> on 
>>>>> a RESOURCEID, add your new row or two, and then remove the filter, 
>>>>> re-sort, 
>>>>> and you will have properly "inserted" a row.
>>>>>
>>>>> Finally, I just downloaded Open Office, and it looks like you can 
>>>>> write Python or Javascript macros.  This is pretty huge (if you're into 
>>>>> that sort of thing) and a simple python or js macro could be used in 
>>>>> place 
>>>>> of the SQL query I made.  Sharing open office macros is probably really 
>>>>> easy, so this would be a great way to collaborate.
>>>>>
>>>>> Good luck!
>>>>>
>>>>>
>>>>> On Sunday, February 14, 2016 at 6:19:35 AM UTC-6, Lucy FJ wrote:
>>>>>>
>>>>>> Hi Adam,
>>>>>>
>>>>>> I have now thoroughly read the documentation on .Arches file and have 
>>>>>> played around with the Microsoft query and have a much better idea of 
>>>>>> what 
>>>>>> I need to do, so there is no need to answer some of the questions I 
>>>>>> asked 
>>>>>> below! 
>>>>>>
>>>>>> It seems that you either need to do a lot of manipulation of the 
>>>>>> excel file or a lot of copy and pasting in the .Arches file which will 
>>>>>> then 
>>>>>> be enormous even for our modest database!  
>>>>>>
>>>>>>
>>>>>>
>>>>>> Lucy
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Sunday, February 14, 2016 at 11:01:23 AM UTC+2, Lucy FJ wrote:
>>>>>>>
>>>>>>> Hi Adam,
>>>>>>>
>>>>>>> Finally I am back testing out creating an .Arches file (but I am not 
>>>>>>> able to load them yet and see the results) and I do have further 
>>>>>>> questions 
>>>>>>> about adding alternative names and several resource classifications. 
>>>>>>>
>>>>>>> Just to be clear that I understood you correctly, the group ID keeps 
>>>>>>> records that are related within a heritage resource together, and the 
>>>>>>> resource ID ensures that all data for a particular resource is kept 
>>>>>>> together? Therefore, each alternative name/nametype would have a 
>>>>>>> separate 
>>>>>>> Group ID from the preferred name/nametype, but the same Resource ID?
>>>>>>>
>>>>>>> Secondly, I am not very familiar with Microsoft query, but it seems 
>>>>>>> that the original Excel file for creating the query from, can only have 
>>>>>>> one 
>>>>>>> 'row' per record or archaeological site in our case. So as you 
>>>>>>> suggested, 
>>>>>>> we would need to put in a new column for each alternative name we have 
>>>>>>> and 
>>>>>>> for each site type classification and of course modify the query to add 
>>>>>>> more statements for the extra names etc. I am looking for a method of 
>>>>>>> reading the Excel file in the way we have set up with all the 
>>>>>>> alternative 
>>>>>>> name being in *one* column, as it would be more efficient for us - 
>>>>>>> sometimes we have 10 alternative names but I don't think Microsoft 
>>>>>>> Query 
>>>>>>> can handle this. 
>>>>>>>
>>>>>>> I don't know if you have any further thoughts on this or know how 
>>>>>>> others have handles this.
>>>>>>> Thank you very much for your help, which is always appreciated! 
>>>>>>> Lucy
>>>>>>>
>>>>>>> On Monday, January 18, 2016 at 7:27:24 PM UTC+2, Adam Cox wrote:
>>>>>>>>
>>>>>>>> Hi Lucy, I'm happy to hear the documentation was helpful.
>>>>>>>>
>>>>>>>> To add more attributes to your .arches file, you'll just need to 
>>>>>>>> add more statements to the SQL query.  First, because the SQL window 
>>>>>>>> is 
>>>>>>>> pretty cramped in Excel, I'd recommend pasting your query into Notepad 
>>>>>>>> ++ 
>>>>>>>> so it's easier to work with.
>>>>>>>>
>>>>>>>> If you look at the two statements that define Name and Name_Type, 
>>>>>>>> you'll see that they both reference "group1".  What you want to do is 
>>>>>>>> make 
>>>>>>>> more groups, so I'd recommend copying and pasting the Name_Type 
>>>>>>>> statement 
>>>>>>>> from the existing query (because it is the most standard) and pasting 
>>>>>>>> it 
>>>>>>>> below, as many times as you need--one time per new attribute that you 
>>>>>>>> need 
>>>>>>>> to add.  Then just make sure to set the group value appropriately: if 
>>>>>>>> you 
>>>>>>>> have an alternate name, you'll have to make a new column for that 
>>>>>>>> name, and 
>>>>>>>> make a column with the correct conceptid (or hard-code the conceptid 
>>>>>>>> into 
>>>>>>>> the new statement), and then put "group3" into each statement. Or, if 
>>>>>>>> you 
>>>>>>>> just have a single column for a standalone attribute like "Site Type", 
>>>>>>>> make 
>>>>>>>> sure it has its own unique group ID.
>>>>>>>>
>>>>>>>> I hope that makes sense.  I have to run now, but if you have more 
>>>>>>>> questions don't hesitate to ask them!
>>>>>>>>
>>>>>>>> Adam
>>>>>>>>
>>>>>>>> On Sunday, January 17, 2016 at 12:17:31 PM UTC-6, Lucinda 
>>>>>>>> Fletcher-Jones wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> <https://lh3.googleusercontent.com/-Qu9c0GYT4ig/VpvaZj1w5yI/AAAAAAAAFXo/aMjV0c5GLJg/s1600/Dabanarti.JPG>
>>>>>>>>>
>>>>>>>>> Hi all,
>>>>>>>>>
>>>>>>>>> I have worked through the example in Adam Cox's excellent guide to 
>>>>>>>>> creating a .Arches file from a Microsoft Excel file which is where we 
>>>>>>>>> have 
>>>>>>>>> all our archaeological site data at present, and it worked very well. 
>>>>>>>>> Now I 
>>>>>>>>> have a question. Our data, as everyone else's is a little more 
>>>>>>>>> complicated 
>>>>>>>>> than the example. How do you deal with a site that has more than one 
>>>>>>>>> name 
>>>>>>>>> and more than one classification etc? for example: Dabanarti with 
>>>>>>>>> three 
>>>>>>>>> names and three site types. 
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> <https://lh3.googleusercontent.com/-Qu9c0GYT4ig/VpvaZj1w5yI/AAAAAAAAFXo/aMjV0c5GLJg/s1600/Dabanarti.JPG>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Thank you very much, Lucy
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> -- 
>>>> -- To post, send email to [email protected]. To unsubscribe, 
>>>> send email to [email protected]. For more information, 
>>>> visit https://groups.google.com/d/forum/archesproject?hl=en
>>>> --- 
>>>> You received this message because you are subscribed to the Google 
>>>> Groups "Arches Project" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send 
>>>> an email to [email protected].
>>>> For more options, visit https://groups.google.com/d/optout.
>>>>
>>>
>>> -- 
>> -- To post, send email to [email protected] <javascript:>. To 
>> unsubscribe, send email to [email protected] <javascript:>. 
>> For more information, visit 
>> https://groups.google.com/d/forum/archesproject?hl=en
>> --- 
>> You received this message because you are subscribed to the Google Groups 
>> "Arches Project" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to [email protected] <javascript:>.
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>

-- 
-- To post, send email to [email protected]. To unsubscribe, send 
email to [email protected]. For more information, 
visit https://groups.google.com/d/forum/archesproject?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"Arches Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to