Android Question Scraping web with html

[3394509365]

Hello, I made a program in VB.net that does web scraping, and pulls down data from a web page.
I would like to do the same thing though with an app in basic4android.
From what I understand there is no library that can intercept the tags and nodes from HTML, but perhaps I should first pull down the page locally in XML and then go there to intercept the tags that interest me. Did I understand correctly or is there any other way to approach the problem?

Do you have any sample programs to start from?

Thanks
Regards

 

[Brian Dean]

I have done what you describe a couple of times. I am not very knowledgeable about HTML syntax or web related topics generally. I found that using simple string functions (like IndexOf() specific tag values) to chop out the section of HTML that contained the data that I needed, followed by Regex searches on the extracted segment worked well and was quite quick and easy to set up. But you might have a more complicated requirement.

 

[3394509365]

Do you have any examples? , I'm trying with parse or httpjob but I'm going crazy

 

[3394509365]

I got

to this point, as per the attachment, but the function

    Dim m As Matcher = Regex.Matcher("table class=""t2""><td class=\""", page)' originale

does not work






this is what I should extract

then cyclically they repeat and change the extraction dates

 

[Jorge M A]

it is difficult to help you without knowing the URL and the structure of the page.
You have to adapt the Regex to the expression you are interested in finding.
you can use the search engine.

Search Results

www.b4x.com

 

[MicroDrie]

There is an jSoup HTML Parser library which can be used for scrapping a HTML page. The advantage of using Jsoup is that there are several examples online how to scrape an HTML page.

Update: Hereby the needed jsoup-1.8.1.jar for the shared library folder.

 

[Brian Dean]

Do you have any examples? , I'm trying with parse or httpjob but I'm going crazy

I separated out the table section that I wanted like this :

        html = html.Replace(Chr(10), "")    ' HTML page with line feeds removed
        Dim start As Int = html.IndexOf($"<table id="performances">"$)
        Dim stop As Int = html.IndexOf2($"</table>"$, start)
        If (stop > start) Then
            html = html.SubString2(start, stop + 8)
            extractList
        End If

The section that I was interested in had a unique start tag (looks like your does too) so you grab everything up to the next </table> tag. Removing the line feeds first helps clean thing up and prepares for Regex searches. Even if you intend to use an XML parser I think that cutting the HTML down to size in this way makes things easier to manage.

I was able to pick out the items that I wanted by a Regex search for <class ...> tags ...

Private Sub extractPerformance(data As String, p As performance)
    Dim pattern As String = $"class=.+?>[^\<]*"$
    Dim m As Matcher = Regex.Matcher(pattern, data)
    Do While m.Find
        item = m.Match
        If item.StartsWith($"class="association""$) Then p.association = extractData(item)
        If item.StartsWith($"class="place""$) Then p.place = extractData(item)
        If item.StartsWith($"class="address""$) Then p.address = extractData(item)
        If item.StartsWith($"class="changes""$) Then changes = extractData(item)
        If item.StartsWith($"class="title""$) Then p.title = changes & " " & extractData(item)
        If item.StartsWith($"class="footnote""$) Then p.footnote = p.footnote & " " & extractData(item)
    Loop
End Sub

' Extract the data content from a RegEx match item
Private Sub extractData(item As String) As String
    Dim result As String
    result = item.SubString(item.IndexOf(">") + 1) "
    Return result
End Sub

This is very low tech but quite flexible; I am not very good at Regex. I don't expect that it will be a perfect fit for your situation, but I hope that it might give you some ideas

 

[3394509365]

it is difficult to help you without knowing the URL and the structure of the page.
You have to adapt the Regex to the expression you are interested in finding.
you can use the search engine.

Search Results

www.b4x.com

in the

Attachment

there is the url

 

[3394509365]

I separated out the table section that I wanted like this :

        html = html.Replace(Chr(10), "")    ' HTML page with line feeds removed
        Dim start As Int = html.IndexOf($"<table id="performances">"$)
        Dim stop As Int = html.IndexOf2($"</table>"$, start)
        If (stop > start) Then
            html = html.SubString2(start, stop + 8)
            extractList
        End If

The section that I was interested in had a unique start tag (looks like your does too) so you grab everything up to the next </table> tag. Removing the line feeds first helps clean thing up and prepares for Regex searches. Even if you intend to use an XML parser I think that cutting the HTML down to size in this way makes things easier to manage.

I was able to pick out the items that I wanted by a Regex search for <class ...> tags ...

Private Sub extractPerformance(data As String, p As performance)
    Dim pattern As String = $"class=.+?>[^\<]*"$
    Dim m As Matcher = Regex.Matcher(pattern, data)
    Do While m.Find
        item = m.Match
        If item.StartsWith($"class="association""$) Then p.association = extractData(item)
        If item.StartsWith($"class="place""$) Then p.place = extractData(item)
        If item.StartsWith($"class="address""$) Then p.address = extractData(item)
        If item.StartsWith($"class="changes""$) Then changes = extractData(item)
        If item.StartsWith($"class="title""$) Then p.title = changes & " " & extractData(item)
        If item.StartsWith($"class="footnote""$) Then p.footnote = p.footnote & " " & extractData(item)
    Loop
End Sub

' Extract the data content from a RegEx match item
Private Sub extractData(item As String) As String
    Dim result As String
    result = item.SubString(item.IndexOf(">") + 1) "
    Return result
End Sub

This is very low tech but quite flexible; I am not very good at Regex. I don't expect that it will be a perfect fit for your situation, but I hope that it might give you some ideas

I'm trying to implement your example, but what is it? p As performance

 

[MicroDrie]

The difference between scrapping an HTML page with Regex or Jsoup is that the latter takes into account the HTML formatting rules. That makes Jsoup library suitable for scraping more complex html pages.

The program below produces a log with the 8 numbers at the table top of the website in both B4A and B4J.

First, in the HTML string variable, the entire web page is pulled. Then the two classes that contains the date (“t1” class) and the numbers themselves (“t2” class). Then, the results of those two classes are used to extract the requested date and numbers. The steps have been commented for clarity.

Sub Process_Globals
    Private fx As JFX
    Private MainForm As Form
    
'    --- Variable for Jsoup
    Private js As jSoup
    Private Extract01 As List
    Private Extract02 As List
    Private Extract03 As List
    Private Extract04 As List
End Sub

Sub AppStart (Form1 As Form, Args() As String)
    MainForm = Form1
'    MainForm.RootPane.LoadLayout("Layout1") 'Load the layout file.
'    MainForm.Show

    Extract01.Initialize
    Extract02.Initialize
    Extract03.Initialize
    Extract04.Initialize
    
    ScrapeTable
    
#IF B4A
    Activity.Finish
#End If

#if B4J
    ExitApplication        ' ends the program
#End If
    
End Sub

Private Sub ScrapeTable

'    --- Load the url
    Dim url As String = "https://www.superenalotto.com/archivio"
    
'    ---  Get the page content
    Dim HTML As String = js.connect(url)
'    Log (HTML)

'    --- Get the date for the numbers    
    Dim FirstDateRow As String = js.getElementsByClass(HTML, "t1").Get(0)

'    --- Get the first (op top row) of the table in t2 class
    Dim FirstTableRow As String = js.getElementsByClass(HTML, "t2").Get(0)

'    --- We need the t1 class for the date    
    Extract01 = js.selectorElementText(FirstDateRow, "a")

'    --- We need only the td rows of the wanted table    
    Extract02 = js.getElementsByClass(FirstTableRow, "ball-24px")
    Extract03 = js.getElementsByClass(FirstTableRow, "jolly-24px")
    Extract04 = js.getElementsByClass(FirstTableRow, "superstar-24px")

'    --- Show results during test periode
'    Log("Extract01: "& Extract01)
'    Dim Count As Int = Extract01.Size
'    Log("Count: "& Count)
'    
'    Log("Extract02: "& Extract02)
'    Dim Count As Int = Extract02.Size
'    Log("Count: "& Count)
    '
'    Log("Extract03: "& Extract03)
'    Dim Count As Int = Extract03.Size
'    Log("Count: "& Count)
    '
'    Log("Extract04: "& Extract04)
'    Dim Count As Int = Extract04.Size
'    Log("Count: "& Count)
    
'    --- Start with the date which is the first line
    Log($"The number from ${Extract01.Get(0)} are: ${CRLF}"$)
    
'    --- We need to scrape the rows and columns
    Dim columns As List
    columns.Initialize
    
'    --- To display a counter at the variables end
    Dim x As Int = 0
    
'    --- Scrape the first 6 numbers with the same class
    For i = 0 To Extract02.Size -1
        columns = js.selectorElementText($"<table>${Extract02.Get(i)}</table>"$, "td")
        
        For j = 0 To columns.Size -1
            x = x + 1
            Log($"Number${x}: ${columns.Get(j)}"$)
        Next
    Next
    
'    --- And scrape the Jolly number with different class
    For i = 0 To Extract03.Size -1
        columns = js.selectorElementText($"<table>${Extract03.Get(i)}</table>"$, "td")
        
        For j = 0 To columns.Size -1
            x = x + 1
            Log($"Jolly Number${x}: ${columns.Get(j)}"$)
        Next
    Next
    
'    --- And scrape the super number with different class 
    For i = 0 To Extract04.Size -1
        columns = js.selectorElementText($"<table>${Extract04.Get(i)}</table>"$, "td")
        
        For j = 0 To columns.Size -1
            x = x + 1
            Log($"Super Number${x}: ${columns.Get(j)}"$)
        Next
    Next
End Sub

 

[3394509365]

I can't load the libraries for B4A, Whi?

 

[MicroDrie]

I can't load the libraries for B4A, Whi?

In my example, I only used the original B4A and B4J jSoup HTML Parser library from 5 years ago.
Although I have not used the cited legacy jsoup.org library version and download link for 1.8.1 jar, the current version at https://jsoup.org/download is the current jsoup-1.13.1.jar libraries available for download.

Update: @3394509365 the five year old library doesn't support all jSoup options. The jSoup org library version 1.13.1 is needed to explore the full jSoup capabilities.

 

[3394509365]

this, I have already downloaded and placed in the B4A library folder but does not see it.

in B4J it works

 

[Brian Dean]

I'm trying to implement your example, but what is it? p As performance

A "performance" is just an object (record) to collect the data values that I am scraping from the HTML. Looking at your example I would first use IndexOf() to pick out the "<table class="t2"> to "</table>" segment, and then scan the "<td . . . . . </td>" chunks. I have not used jSoup; I must say that it looks pretty good, but may be over-powered for what you need, and you have the overhead of learning a new library. But this is a personal preference. You should look for what best suits your previous coding experience.

 

[3394509365]

while I solve the problem with B4A I saw in b4j that the library works very well, excellent job.

I attach the extract of the page, as the structure of the page repeats the same many times but so I extract it only once,

I need to do more laps.

eg the draws from January with the numbers and dates, at the moment it is only the last one.

 

[MicroDrie]

while I solve the problem with B4A I saw in b4j that the library works very well, excellent job.
I attach the extract of the page, as the structure of the page repeats the same many times but so I extract it only once,
I need to do more laps.
eg the draws from January with the numbers and dates, at the moment it is only the last one.

Because you give a picture and mark only the last result of the lottery I take only the first one

'    --- Get the date for the numbers    
    Dim FirstDateRow As String = js.getElementsByClass(HTML, "t1").Get(0)

you have two choices:

1. Drop the .Get(0) to get all the possibilities or
2. Replace the zero by the result of a for next loop whereby you must define the range like the last year ( 0 to 11

 

[3394509365]

I took a few steps forward until I resolved the library issue.

Sub InterpretaTabella(WWW As String)
    Dim TabellaHTML As String
    TabellaHTML=WWW.SubString(WWW.ToUpperCase.IndexOf($"<TABLE CLASS="T2">"$)) 'funziona la data
    TabellaHTML=TabellaHTML.SubString2(0,TabellaHTML.ToUpperCase.IndexOf("</TR"))'originale
    Dim Righe() As String = Regex.Split("</td>",TabellaHTML)'originale
  
    For Rg=0 To Righe.Length-1
        Dim celle() As String = Regex.Split("</td>",Righe(Rg))
        For Col=0 To celle.Length-1
            Dim Cella As String
            Cella=TogliTag(celle(Col))
            'Log("Cella (" & Rg & "," & Col & ") :" & Cella)
            Log (Cella.Trim)
        Next
    Next
End Sub

Sub TogliTag(cella As String) As String
    Dim Ret As String =""
    Dim Tag As Boolean = False
 
    For i=0 To cella.Length-1
        If cella.SubString2(i,i+1)="<" Then Tag=True
        If Not(Tag) Then Ret=Ret & cella.SubString2(i,i+1)
        If cella.SubString2(i,i+1)=">" Then Tag=False
    Next

    Return Ret
End Sub

now I can extract the first column of numbers, but as you can see in the attachment, the cycle repeats itself many times, I have to understand how to do it.

 

[MicroDrie]

Have you tried in B4A to resync the libraries and reload the program code to fix the problem? I take a look to your text file processing.

 

[MicroDrie]

As I wrote before the jSoup library is designed to deal with the complex HTML structure. Here is the code to read your text file with jSoup:

'    --- Load the raw text file
    Dim RawHTML As String = ""
    Dim read1 As String="miotesto.txt"
    RawHTML = File.ReadString(File.DirAssets,read1)
'    Log("Raw RawHTML: " & RawHTML)  'displays the content of file
       
'    --- Remove empty lines and reformat the layout
    Dim HTML As String = js.parse_HTML(RawHTML)
    Log("Clean HTML: " & HTML)

This code reformat the HTML structure so maybe you can use it to simplify your regex approach. Feel fry to use it, meanwhile I go study inline Java.

 

[3394509365]

Have you tried in B4A to resync the libraries and reload the program code to fix the problem? I take a look to your text file processing.

yes, i tried,

??