After the successful transfer of my articles from Foxite, I finally converted another bunch of articles from my old blog software. The process itself was quite easy: query database, convert records to XML and write out files.

The interesting part is that the original data was stored in a MySQL database. So, I had to install the MyODBC driver to gain access to it and the rest is pure FoxPro. Have a look at the code:

Migration: blogMachine to DasBlog

    Set Safety Off    Local lnHandle, lnResult, lcAlias, lcSqlStatement, ;         laSql[1], ltDate, lcXml, lcFile, lcContent    m.lnHandle = 0
    m.lnResult = 0
    m.lcAlias = "blogmachine"    m.lcSqlStatement = ""    m.lcXml = ""    m.lcFile = ""    m.lcContent = ""        m.lnHandle = SQLStringConnect("DRIVER={MySQL ODBC 3.51 Driver};" + ;
        ";PORT=3306;" + ;
        "DATABASE=blog;OPTION=19035;UID=joki;PWD=topsecret", .T.)
    If m.lnHandle > 0
        DECLARE INTEGER CoCreateGuid IN Ole32.dll STRING @lcGUIDStruc

        Text To m.lcSqlStatement NoShow Flags 1 Pretext 1+2+4+8
            Select post.title As p_title, post.summary As p_summary,
       As p_date, cat.cat_name As c_name
            From bmc_posts post
            Left Outer Join bmc_cats cat On =
            Order By        EndText        m.lnResult = SQLExec(m.lnHandle, m.lcSqlStatement, m.lcAlias, laSql)
        If m.lnResult > 0 .And. Used(m.lcAlias)
                m.ltDate = Datetime(1970,1,1,0,0,0) + Nvl(Cast(p_date As Integer), 0)
                m.lcFile = GetWordNum(Ttoc(Ttod(m.ltDate),3),1,"T") + ".dayentry.xml"                Text To m.lcXml TextMerge NoShow Flags 1 Pretext 4
<?xml version="1.0" encoding="utf-8"?>
<DayEntry xmlns:xsi=""
<Content><<Strtran(Alltrim(p_summary), Chr(13)+Chr(10), "&"+"lt;br"+"&"+"gt;"+Chr(13)+Chr(10))>></Content>
<EntryId><<Lower(Chrtran(GetGuid(), "{}", ""))>></EntryId>
<Description />
<Attachments />
<Crossposts />

                If File(m.lcFile)
                    m.lcXml = StrExtract(m.lcXml, "<Entry>", "</Entry>", 1, 4)
                    m.lcContent = FileToStr(m.lcFile)
                    m.lcXml = Strtran(m.lcContent, "</Entries>", m.lcXml + Chr(13) + Chr(10) + "</Entries>", 1, 1, 1)
                ? Alltrim(p_Title) + " " + Transform(m.ltDate)
                StrToFile(m.lcXml, m.lcFile)
        Else            Assert .F.            ? AError(m.laSql)        EndIf        SQLDisconnect(m.lnHandle)    EndIfPROCEDURE GetGuid() As String    cStrucGUID=SPACE(16)
    IF CoCreateGuid(@cStrucGUID) # 0
     RETURN ""    ENDIF    IF StringFromGUID2(cStrucGUID,@cGuid,nSize) = 0
There are some interesting parts in this conversion of data:
  • First, the MySQL field names are resevered key words in VFP, so that they have to be renamed before usage.
  • Second, datetime conversion from Unix datetime to 'normal' datetime. This is quite simple:
m.ltDate = Datetime(1970,1,1,0,0,0) + Nvl(Cast(p_date As Integer), 0)
  • Third, datetime string conversion between VFP and .NET. Herefore, we simply use VFP Ttoc() function with '3' as second parameter:

This provides a proper character string with XML DateTime format. See VFP's help file for details.

  • And fourth, generation of a Global Unique IDentifier (GUID) in VFP. This is done with two Win32 API calls. Also nothing extraordinary.

At the moment, this conversion gives me some more categories and increases the number of article to 85. But as you might see on the 'Archive' box on the right side, there is still a huge gap between October 2006 and January 2008. There are more articles to come, but this requires another conversion. The missing entries are part of my AfpWiki database which runs on Microsoft SQL Server. So, stay tuned, you'll get a full package within these days.

Sincerely, JoKi