As far as I understood the check with then is mandatory.

So what is the semantic of having to declare a local to avoid compiling error (object could be void)?

detachable dog: DOG

take_a_walk
    do
        check
            attached_dog: attached dog as l_dog
        then
            leash (l_dog) -- why not just leash (dog)?
            walk
        end
    end     

I'd like to create a class with both an instance_free (ensure class) feature and a deferred one.

why is that I can't call an instance_free feature in a deferred class but only on a descendant?

Don't understand the logic of that... is there for compiling reasons? or any sense I don't see? Where am I wrong with my logic?

Didn't find the answer on this doc, is the rescue clause called when redefining a feature?

Class A

class A

feature -- 

    process
        do
            do_stuff
        rescue
            on_fail
        end
end -- class

Class B

class B

inherit
    A
        redefine
            process
        end

feature -- 

    process
        do
            do_other_stuff -- do I have to call the rescue too?
        end
end -- class

I'm looking for a way to call Precursor on another feature. The context is that I redefine a routine calling another heir's feature that calls mine, and I'd like to call it explicitly. Is there a way to do something such as in java Precursor.feature_a.

If not, the only alternative I find is to write a feature_a_implementation and call it from redefined feature. Is there a particular reason not to have this mechanism? The consequences of doing that would be to define 2 times the contract of feature_a into feature_a and feature_a_implementation

Is there a way to get the project name and target at execution time or is it not stored into executable?

As mentioned on this question, there is no way to define a constant which I can redefine into a descendant.

In many of my cases, I'd like to have a constant which I can redefine. The alternatives I see to avoid a creation on each consultation which doesn't make sense would be

That's not doable

class A

feature -- Access

    Default_value: STRING = "some A value"

end -- Class


class B

inherit
    B
        redefine
            Default_value
        end

feature -- Access

    Default_value: STRING = "some B value"

end -- Class

Alternative with once instance_free

class A

feature -- Access

    Default_value: STRING
        once
            Result := "some A value"
        ensure
            instance_free: class
        end

end -- Class


class B

inherit
    B
        redefine
            Default_value
        end

feature -- Access

    Default_value: STRING
        once
            Result := "some B value"
        ensure
            instance_free: class
        end

end -- Class

As far as I understand, the once wouldn't be created with B value as the A class value would be taken

Alternative with attribute

class A

feature -- Access

    Default_value: STRING
        attribute
            Result := "some A value"
        ensure
            instance_free: class
        end

end -- Class


class B

inherit
    B
        redefine
            Default_value
        end

feature -- Access

    Default_value: STRING
        attribute
            Result := "some B value"
        ensure
            instance_free: class
        end

end -- Class

Would it be the only and good practice to do it?

I'd like to do grouping of regular expressions with eiffel. How do I do something like

l_reg.compile ("^([0-9]{3}) (rabbit[0-9]).*")
l_groups := l_reg.groups ("123 rabbit1")
my_first_rabbit := l_groups.at (2)

Didn't find any example on groups, LX_DFA_REGULAR_EXPRESSION class and other googlings

Id like to uncompress a file with bzip2 myFile.bz2 which class do I have to use for that?

I tried to find it into base kernel documentation which made the most sense for me and didn't find it

How do I concatenate a percent % string to a string?

as either do not work

s := "%%" + s + "%" 

I know its a special character but don't afford it neither do I find the doc...

UPDATE

The answer was one of my attempts, but it could stay an issue of the compiler/editor. After putting the example into the same feature, compiling, and replacing the old code the compiler did not complain anymore, but still stays as syntax coloration strange

As per this documentation, I'm trying to constrain a Generic parameter to 2 classes and need a default_create creation procedure call in my implementation.

I'm getting a Syntax error from compiler on the select clause. Why is that so?

I thought first that calling default_create will call the G one which could be a merge/renaming/etc. but seems that on that level I have to call either one of both or both renamed which makes sense for me. But why does this syntax fail?

The context and my code is following:

deferred class
    IDENTIFIABLE_CASH[G -> {IDENTIFIABLE[ANY],
                            DB_ENTITY select default_create end} create default_create end]

inherit
    DB_SERVICE_CASH[G]


feature -- Access

    cash_from_identifier (an_identifier: like {IDENTIFIABLE_DB_ENTITY[ANY]}.identifier): G
            -- Returns firt entity found like given an_identifier (like operator is used)
        local
            l_qry: COMPOSED_SQL_QUERY
        do
            across
                cash is l_item
            loop
                if l_item.identifier.is_equal(an_identifier) then
                    Result := l_item
                end
            end
            if not attached Result then
                create l_qry.make (item_prototype)
                l_qry.add_constraint (create {FIELD_CONSTRAINT}.make (Void, identifier_db_column_name, {CONSTRAINT}.Like_operator, an_identifier))
                load_entities (l_qry)
                if items.count > 0 then
                    Result := items.first
                else
                    create Result
                    check
                        item_not_found: False
                    end
                end

                cash.extend (Result)
            end
        end

Am looking for typical

feature {NONE} -- Initialization (semantic)
feature -- Status Settings (semantic)
.....

Where can I find that list with semantic?

Didn't find it into ECMA

Got a Catcall trying to compare 2 types, how could I do that avoiding passing through another not dedicated method (like to string, class_id or stuff like that)?

SIT_UTIL

    class_name_lowercase (a_string: STRING): STRING
            -- a copy lowercased and pruned from preceding '!'
        do
            Result := a_string
            if Result.index_of('!', 1) = 1 then
                Result := Result.substring (2, Result.count)
                Result.to_lower
            else
                Result := Result.as_lower
            end
        ensure
            instance_free: class
        end

CLIENT_CLASS

    relationship_from_secondary_type_equal (a_type: like relationships.item.secondary_type): detachable like relationships.item
            -- Returns first instance of found relationship secondary type which equals given one
        do
            across
                relationships as l_rel
            until
                Result /= Void
            loop
--              if attached (a_type / l_rel.item.secondary_type) then -- Don't want conformance but equality

--              if attached (a_type.is_equal (l_rel.item.secondary_type)) then -- tried but as is_equal needs a like Current => Catcall
--              if attached (a_type.equal (a_type, l_rel.item.secondary_type)) then -- Catcall because b signature is like a
                if {SIT_UTIL}.class_name_lowercase (a_type).is_equal({SIT_UTIL}.class_name_lowercase (l_rel.item.secondary_type)) then
                    Result := l_rel.item
                end
            end
            check
                not_found_relationship: Result /= Void
            end
        end

I'm trying to do something as

work (a_father: FATHER)
    do
        if a_father.conforms_to ({DEVELOPER}) then
           a_father.code
        else
           a_father.change_job ({DEVELOPER})
        end
    end

the compilation works, but in my implementation @runtime it doesn't pass. What did I mistype?

Is there a way (I'm sure there is out of runtime check...) to specify that a parameter or a variable in general conforms to multiple types? to avoid doing something such as

work (a_printer: PRINTER; a_scanner: SCANNER)
    do
      a_printer.print
      a_scanner.scan
        -- OR without second parameter
      if attached {SCANNER} a_printer as l_scanner then
         l_scanner.scan
      else
         throw RuntimeError
      end
    end

I attempted to install the Traffic library for the book "A Touch of Class" but was not able to compile the examples. EiffelStudio complained about a base2.ecf file. Error code: VD00

General configuration parsing error. What to do: fix the configuration file.

Could not open file: \base2.ecf $EIFFELBASE2\base2.ecf

base2.ecf is found in where Windows installed Eiffel: C:\Program Files\Eiffel Software\EiffelStudio 18.11 GPL\unstable\library\base2.ecf The offending line of code seems to be this in the Traffic configuration files:

I have attempted to change this configuration file pointing directly to where the file is, using the actual path, to no avail.

I'm really stuck in how to pass an argument to a second level of an Eiffel agent call. Hope my example is enough self explaining:

main_call
    do
        b (agent d(?, ?))
    end

b (a_proc: PROCEDURE[STRING])
    local
        l_something_else: INTEGER
    do
        l_something_else := 1
--        z("a_line", a_proc(l_something_else)) -- How can I pass at this stade l_something_else to the d procedure!!!
    end

d (line: STRING; something_else: INTEGER)
    do
--        do_some_stuff_with_line (line, something_else)
        do_nothing
    end

z (a_line: STRING; a_proc: PROCEDURE[STRING])
    do
        a_proc.call([a_line])
    end

I would have liked to be able to do something like

z("a_line", a_proc(?, something_else))

But its not possible, as I try with the agent keyword, the a_proc argument is not recognized!

So what would be the syntax? I even tried to add an argument to a_proc with a_proc.set_operands but am lost with the OPEN_ARGS class

Implementation case

If you need a goal... just imagine I'd like to have 2 different implementations of the d function, and in my case its using UT_CSV_HANDLER that I'd like to have 2 different functions for each line of a CVS

The following code gives a Non-compatible actual argument in feature call

Feature: import_from_csv_impl
Called feature: import_from_csv_impl (a_rest_request: REST_REQUEST; a_procedure: PROCEDURE [DS_ARRAYED_LIST [STRING_8], INTEGER_64, STRING_8]): [detachable like items] detachable SIT_LINKED_LIST [MEASURING_POINT] from MEASURING_POI...
Argument name: ia_name
Argument position: 2
Formal argument type: STRING_8
Actual argument type: INTEGER_64
Line: 258
                ia_procedure.call (ia_measuring_point_id, ia_name)
->            end (?, l_measuring_point_id, l_s)
            l_csv_handler.read_file (l_is, l_partially_closed)

The complete example:

-- Main call
import_from_abb_csv (a_rest_request: REST_REQUEST): detachable like items 
    do
        Result := import_from_csv_impl (a_rest_request, agent impl_for_each_csv_line_import_from_abb_csv)
    end

-- Second call
import_from_csv_impl (a_rest_request: REST_REQUEST; a_procedure: PROCEDURE[DS_ARRAYED_LIST [STRING_8], INTEGER_64, STRING]): detachable like items
    local
        l_csv_handler: UT_CSV_HANDLER
        l_is: KL_STRING_INPUT_STREAM
        l_measuring_point_id: INTEGER_64
        l_s: STRING
        l_partially_closed: PROCEDURE[DS_ARRAYED_LIST[STRING]]
    do
        l_s := "whatever"
        l_measuring_point_id := 12
        create l_csv_handler.make_with_separator (',')

        l_partially_closed := agent (i_al: DS_ARRAYED_LIST[STRING]; ia_measuring_point_id: INTEGER_64; ia_name: STRING; ia_procedure: PROCEDURE[INTEGER_64, STRING])
            do
                ia_procedure.call (ia_measuring_point_id, ia_name)
            end (?, l_measuring_point_id, l_s)

        l_csv_handler.read_file (l_is, l_partially_closed)

    end

-- end call
impl_for_each_csv_line_import_from_abb_csv (a_csv_line: DS_ARRAYED_LIST [STRING_8]; a_measuring_point_id: INTEGER_64; l_cu_name: STRING)
    do
        -- do_my_business
    end


-- for information signature of read_file is:
--     read_file (a_file: KI_TEXT_INPUT_STREAM; a_action: PROCEDURE [DS_ARRAYED_LIST [STRING]])

General

I don't understand why there are like 2 generating_type kinds, the one with exclamation point, and the other not.

My case

I have a function returning a DB_SERVICE[G] from a given instance from a collection of db_services so that from another service I can retrieve it and call it. When I pass an instance generating_type to that function, the conforms_to returns True, as when I give ({ENUMERATE}).generating_type it doesn't.

Why is that so?

-- Humanly unreadable
if attached {ENUMERATE_DB_SERVICE} {SIT_ENVIRONMENT}.app_instance.db_service_from_entity_type (item_prototype.charge_unit_relationship.secondary_prototype.an_index_relationship.secondary_prototype.generating_type) as l_enum_dbs then

-- Humanly readable but not working
if attached {ENUMERATE_DB_SERVICE} {SIT_ENVIRONMENT}.app_instance.db_service_from_entity_type (({ENUMERATE}).generating_type) as l_enum_dbs then

My function

db_service_from_entity_type (an_entity_type: TYPE[detachable DB_ENTITY]): detachable like db_services.item
    do
        across
            db_services as l_dbs
        until
            Result /= Void
        loop
            if l_dbs.item.item_prototype.generating_type.conforms_to (an_entity_type) then
                Result := l_dbs.item
            end
        end
    ensure
        service_found: attached Result
    end

Edit (20190405-11:26 UTC)

As the screenshot shows, giving {ENUMERATE} instead of ({ENUMERATE}).generating_type doesn't work either

Seems I'm wrong creating a datetime from string using this method, any idea why? The precondition violates on date_time_valid: date_time_valid (s, code)

create l_date_time.make_from_string ("2019-03-28T15:28:02Z", "yyyy-mm-ddThh:mi:ssZ")

If I'd like to add a DATE_TIME to a logger for example I'd like to avoid creating a local variable for that. Is there any similar statement as

create {DATE_TIME}.make_now.formatted_out ("yyyy-mm-dd hh:mi:ss")

which is not possible?

As I have the choice on finalizing a project to keep assertions, I'd think that assertion violation would be shown or displayed on console (hopefully on stderr on linux). What would be the proper way to do such, like seems that I have to add it or enable something to get it...

Some of my tests needs something like

{EXECUTION_ENVIRONMENT}.arguments.separate_character_option_value ('l')

to be attached (not Void), how do I set commandline arguments to eiffel autotest?

It's not the first time I'm trying to inherit from a class and convert an object to it, in this case extend functionalities of HTTP_CLIENT_RESPONSE with valid_http_response => response.status = 200, valid_http_json_body => validate json content, etc...

For that, I'd like to inherit from HTTP_CLIENT_RESPONSE and add features and create my custom DB_ENTITY_HTTP_CLIENT_RESPONSE from an HTTP_CLIENT_RESPONSE

test_case
    local
        l_http_client_r: HTTP_CLIENT_RESPONSE
        l_db_entity_http_client_r: DB_ENTITY_HTTP_CLIENT_RESPONSE
    do
        l_http_client_r := execute_get("someURL") -- returns an HTTP_CLIENT_RESPONSE
        l_db_entity_http_client_r := l_http_client_r
        assert("valid response", l_db_entity_http_client_r.valid_response)
    end

it seems I've difficulties having the internal properties set... what would be the best way to do this? I had the same case trying to create a WATT class inheriting from NATURAL_32 which is an expanded.

In my strategy, I tried to call in creator

• a parent creation procedure
• then call a deep_copy

Here is the rest of my class attempt:

class
    DB_ENTITY_HTTP_CLIENT_RESPONSE

inherit
    HTTP_CLIENT_RESPONSE

create
    make_from_http_client_response

convert
    make_from_http_client_response ({HTTP_CLIENT_RESPONSE})

feature -- Initialization

    make_from_http_client_response (a_client_response: HTTP_CLIENT_RESPONSE)
        do
            make (a_client_response.url)
            deep_copy (a_client_response)
        end

feature -- Status report

    valid_response: BOOLEAN
        do
            Result := status = 200
        end

The only way I found working for now is by setting all attributes to other which is the semantic of deep_copy normally...

make_from_http_client_response (a_client_response: HTTP_CLIENT_RESPONSE)
    do
        make (a_client_response.url)
        set_body (a_client_response.body)
        set_http_version (a_client_response.http_version)
        set_error_occurred (a_client_response.error_occurred)
        set_error_message (a_client_response.error_message)
        set_raw_header (a_client_response.raw_header)
        set_status_line (a_client_response.status_line)
        ... I surely forgot something...
    end

across
    collection as l_item
until 
    Result /= Void
loop
    if l_item.item.name.is_equal ("foo") then
        Result := l_item.item
    end
end

is there a way and if so which one to do something like

collection.do_if (agent ...)

an example of use can be:

search_item_with_id (an_id: INTEGER)
        -- Moves items cursor if not found is_after
    local
        l_found: BOOLEAN
    do
        from
            items.start
        until
            items.after or l_found
        loop
            l_found := items.item.primary_key = an_id
            if not l_found then
                items.forth
            end
        end
    ensure
        cursor_on_element_if_found: not items.after implies items.item.primary_key = an_id
    end

Eiffel

Seems I'm getting an INTEGER_32_REF from a BIGINT from ODBC with Eiffel DB_SELECTION.cursor. Sometimes I seem to have the adecuate value, but sometimes not...

946383958 => good INTEGER_32_REF (id->4 of following table)
9475984837 => 886050245 WRONG INTEGER_32_REF (id->4 of following table)

646383958 => good INTEGER_32_REF (id->3 of following table)
6475984837 => -2113949755 WRONG INTEGER_32_REF (id->4 of following table)

...

It seems to behave like not treating the BIGINT and giving the INTEGER_32_REF value directly interpreted from the DB value, typical case passing over the 2^32

Postgresql

Having a table with following data as electricity_amount BIGINT NOT NULL DEFAULT 0

id |   a_date   |  amount   | electricity_amount | consumption_sector_id | electricity_rate_id
----+------------+-----------+--------------------+-----------------------+---------------------
1 | 2019-01-01 | 746383958 |         7475984837 |                     1 |                   1
2 | 2019-02-01 | 846383958 |         8475984837 |                     1 |                   1
3 | 2019-03-01 | 646383958 |         6475984837 |                     1 |                   1
4 | 2019-04-01 | 946383958 |         9475984837 |                     1 |                   1

I don't understand why it is so.

• If I say attached {G} it works
• If I say nothing which would be the expected behavior for me calling_entity: detachable RELATED_DB_ENTITY should be conform (once attached) to G which is -> DB_ENTITY
• If I say DB_ENTITY it doesn't
• If I say RELATED_DB_ENTITY either does it pass

Why do I have to specify {G}???

SIT_HANDLER

class
    SIT_HANDLER[G -> DB_ENTITY create default_create, make_from_db_service, make_from_json end]

feature --

    some_feature
        do
            if attached {G} l_rest_request.calling_entity as l_calling_entity then
                db_service.set_item_prototype (l_calling_entity) -- Complains here!!!!!!!!!!!!
                db_service.load_with_relationships (l_rest_request)
            ...
        end
end -- class

REST_REQUEST

class
    REST_REQUEST

feature -- Access
    calling_entity: detachable RELATED_DB_ENTITY -- RELATED_DB_ENTITY inherits DB_ENTITY
...
end -- class

DB_SERVICE

class
    DB_SERVICE [G -> DB_ENTITY create default_create, make_from_db_service, make_from_json end]

feature -- Status setting

    item_prototype: G

    set_item_prototype (v: like item_prototype)
        do
            item_prototype := v
        ensure
            item_prototype = v
        end

...
end -- class

How to set a constant refering to another class constant in Eiffel?

Something like that doesn't compile unfortunately

Default_log_level: like {MY_LOGGER}.log_level = {MY_LOGGER}.Log_level_info

Trying to transform a real_32 to real_64, I'm getting

real_32: 61.55
real_64: 61.54999923706055 

Am I wrong with the to_double function?

Didn't find a better way to format a multiline string than this way... seems complicated. What would be the best way to format this type of code?

l_qry := "SELECT%
    %    * %
    %FROM %
    %    enumerate %
    %WHERE %
    %    " + {like item_prototype}.Primary_key_db_column_name + " = " + l_category_id + " %
    %UNION %
    %   SELECT %
    %       e.* %
    %   FROM %
    %       enumerate e %
    %INNER JOIN  %
    %   enumerates_leaves s ON s." + {like item_prototype}.Primary_key_db_column_name + " = e." + {like item_prototype}.Category_db_column_name + " %
    %) SELECT * FROM enumerates_leaves WHERE enumerates_leaves." + {like item_prototype}.Category_db_column_name + " IS NOT NULL;"

I was wondering if the across structure uses an own cursor or a separated one? does it ensure that cursor hasn't moved and if so how can it be expressed for other examples?